19 Seminario Internacional de Alta Tecnologia

19° Seminário Internacional de Alta Tecnologia

Inovações Tecnológicas no Desenvolvimento do Produto

Editor

Prof. Dr.-Ing. Klaus Schützer

Lab. de Sistemas Computacionais para Projeto e Manufatura

Faculdade de Engenharia, Arquitetura e Urbanismo

09 de Outubro de 2014

Universidade Metodista de Piracicaba

Teatro UNIMEP - Campus Taquaral, Piracicaba, SP

Lab. de Sistemas Computacionais para Projeto e Manufatura Prof. Dr.-Ing. Klaus Schützer FEAU - UNIMEP

Arte e Edição:

André Fagionatto de Castro Bruna Sakamoto

Dhiogenes dos Santos Sousa

Seminário Internacional de Alta Tecnologia (19. : 2014 : Piracicaba, SP)

S471a Anais do 19. Seminário Internacional de Alta Tecnologia, Piracicaba, SP, Brasil, 09 out., 2014 / editor Klaus Schützer. – Piracicaba: UNIMEP, 2014. 234 p. : il. ; 30 cm.

Tema central: Inovações tecnológicas no desenvolvimento do produto

ISSN 2175-9960

1. Seminários (Estudo) - Tecnologia de ponta. I. Schützer, Klaus. II. Título.

CDU – 62

I


A Universidade Metodista de Piracicaba (UNIMEP) é herdeira de uma tradição de mais de

265 anos em educação, iniciada em 1748, com a fundação da Kingswood School, primeira

escola Metodista na Inglaterra. Desta semente inicial, a educação metodista se expandiu,

alcançando atualmente mais de 775 instituições presentes em 70 países, em todos os

continentes. Inúmeras delas são universidades renomadas, como a Emory University

(Geórgia, EUA), Duke University (Carolina do Norte, EUA), American University (Washington

DC, EUA), Southern Methodist University (Texas, EUA), Roehampton University (Londres,

Reino Unido), Africa University (Zimbabwe), Aoyama Gakuin University (Tókio, Japão),

Hiroshima Jogakuin University (Hiroshima, Japão), Yonsei University e Ewha Womans

University (Seoul, Coréia do Sul), Boston University (Massachusetts, EUA) e a Southwestern

University (Texas, EUA), dentre outras. Destas duas últimas, saíram cinco prêmios Nobel,

quatro na área de medicina e um na área de química. Entre os ex-alunos, destacam-se

personalidades como Martin Luther King, responsável pela revolução dos Direitos Humanos

nos Estados Unidos, na década de 1960.

No Brasil, a educação metodista está completando 133 anos, contando atualmente com duas

universidades: UNIMEP e Universidade Metodista de São Paulo - UMESP; dois centros

universitários, Instituições com cursos da Educação Infantil à Superior e mais de vinte que

atuam desde creches até o Ensino Básico e Fundamental; três Faculdades de Teologia e seis

Seminários Teológicos Regionais, que agregam um universo de aproximadamente 62 mil

estudantes. No total são 52 instituições associadas à Rede Metodista de Educação no Brasil,

além de unidades especiais, como a Escola de Música de Piracicaba “Maestro Ernst Mahle”.

A Universidade Metodista de Piracicaba - UNIMEP, com aproximadamente 11.200

estudantes distribuídos em seus quatro campi, localizados nas cidades de Piracicaba, Santa

Bárbara d’Oeste e Lins, matriculados em 53 cursos de graduação, incluindo os cursos

superiores de tecnologia, 20 especializações (26 turmas), 7 mestrados e 4 doutorados, tem

a certeza de estar oferecendo uma educação diferenciada e de qualidade, baseada nos

princípios de sua Política Acadêmica, que tem como premissa a construção da cidadania

como patrimônio da sociedade.

A UNIMEP possui inúmeras parcerias com universidades de renome das Américas do Sul,

Central e do Norte, mantendo, também, parcerias e projetos com a Europa, Ásia e alguns

países da África. Dentre as principais parceiras destaca-se a Darmstadt University of

Technology na Alemanha, com a qual a UNIMEP desenvolve projetos de pesquisa,

intercâmbio de professores e estudantes, além da organização conjunta deste Seminário

Internacional.

II

Além da Darmstadt University of Technology, a UNIMEP mantém convênios com Marietta

College, University of Evansville, Universidad del Centro Educativo Latinoamericano

(Argentina), Universidad Madero (México), Universidad del Sevilla (Espanha), Technische

Universitat Berlin (TUB), Fraunhofer Institut für Produktionsanlagen und Konstruktionstechnik

(Fraunhofer IPK) e Nagasaki Wesleyan University (Japão). A Assessoria para Assuntos

Internacionais também realiza programas para aprendizado de línguas estrangeiras, como

inglês e espanhol. Não só envia alunos para o exterior, mas também recebe muitos alunos

para realizar estudos de curta ou longa duração, na graduação ou pós-graduação.

Mantendo a tradição de inovação e participação na comunidade, professores e alunos da

UNIMEP têm se destacado nas pesquisas e publicações, no intenso aproveitamento dos mais

de 100 laboratórios disponíveis, na prestação de serviços a empresas e à comunidade, no

desenvolvimento de um ambiente de estudos que favorece a convivência e o trabalho

conjunto, e no incentivo à busca das mais variadas oportunidades profissionais, através de

estágios supervisionados e convênios com indústrias, órgãos públicos e universidades no

Brasil e no exterior.


A Faculdade de Engenharia, Arquitetura e Urbanismo (FEAU)/UNIMEP, localizada em Santa

Bárbara d’Oeste, oferece sete cursos de Engenharia (Produção, Controle e Automação,

Mecânica, Alimentos, Química, Elétrica e Civil), Bacharelado em Química, Arquitetura e

Urbanismo e Tecnologia em Processos Químicos. Além disso, oferece um Programa de Pós-

graduação “stricto sensu” (Mestrado e Doutorado) em Engenharia de Produção.

Atualmente a FEAU tem cerca de 2.200 alunos nos cursos de graduação e pós-graduação.

Conta também com um corpo docente de 70 professores, sendo cerca de 55% em regime de

dedicação integral ou parcial. Cerca de 80% do corpo docente é titulado, sendo que 32

professores já concluíram o doutorado no Brasil ou no exterior. Este corpo docente tem

possibilitado o desenvolvimento de diversos projetos de pesquisa com financiamento de

órgãos governamentais (FAPESP, FINEP, CNPq, Capes, etc.), da iniciativa privada (Sandvik,

Indústrias Romi, Siemens PLM, Hexagon Metrology, Eletrocast, IBM entre outros) ou ainda

de organismos externos (DAAD, DFG, etc.).

A FEAU conta com 36 laboratórios para ensino e/ou pesquisa, entre eles um dos mais

modernos laboratórios para ensino de CAD e CAM dentre as universidades brasileiras, além

de duas Salas Ambiente que representam uma nova proposta para o uso da informática no

ensino da engenharia. Alguns de seus laboratórios de pesquisa têm se destacado no Brasil

III

e no exterior pelo trabalho desenvolvido, como nas áreas de sistemas CAD/CAM e PLM,

usinagem com altíssima velocidade, metrologia, materiais carbonosos, dentre outros.

A FEAU tem uma inserção bastante grande junto aos órgãos públicos, ONGs e no parque

industrial e de serviços regionais, que compreende além da região de Piracicaba, a região

Metropolitana de Campinas, através de convênios e projetos. A Incubadora de Empresas de

Santa Bárbara d’Oeste é gerida pela FEAU através de um de seus professores. Além disso,

a Faculdade mantém um forte contato internacional com Universidades e Instituições de

Pesquisa principalmente na Alemanha, Espanha, Bélgica, EUA e Argentina, através de

programas de intercâmbio entre professores, pesquisadores e alunos de graduação e pós-

graduação.

Laboratório de Sistemas Computacionais para Projeto e Manufatura

O Laboratório de Sistemas Computacionais para Projeto e Manufatura (SCPM) é um dos

mais de 30 laboratórios da Faculdade de Engenharia, Arquitetura e Urbanismo (FEAU) da

Universidade Metodista de Piracicaba. Na sua maioria, esses laboratórios estão voltados

primordialmente ao ensino, possibilitando aos estudantes um primeiro contato com a

realidade que enfrentarão no mercado de trabalho.

O SCPM, no entanto, foi criado com foco na pesquisa, residindo aí o seu diferencial, ou seja,

sua finalidade primeira é possibilitar a iniciação científica, através de projetos a serem

desenvolvidos pelos estudantes sob supervisão de professores. Esse é o papel que vem

desempenhando ao longo dos seus 19 anos de existência, sem descuidar da preservação da

indissociabilidade das duas outras colunas de sustentação de uma universidade, ou seja, o

ensino e a extensão.

Como uma das primeiras atividades, o Laboratório instalou os equipamentos de informática

recebidos através de dois projetos, o KIT #123 - FBaseDsgn, financiado pela Comissão

Europeia, e o projeto para implantação de infraestrutura, financiado pelo Deutsche

Ausgleichsbank. Em torno desse trabalho aglutinou-se um grupo de alunos de graduação e

pós-graduação que ajudou no planejamento e organização da primeira versão do que se

tornou o Seminário Internacional de Alta Tecnologia. O primeiro evento, em 1996, introduziu

no Brasil a temática da “Usinagem com Altíssima Velocidade de Corte”, que hoje é tema de

pesquisa em várias universidades e aplicada em diversas empresas.

Este grupo de pesquisa criou também o Núcleo para Projeto e Manufatura Integrados (NPMI),

incluído no Cadastro Nacional de Grupos de Pesquisa do CNPq desde 1995, e que oferece

IV

a interface para integração de outros professores e pesquisadores aos trabalhos

desenvolvidos no SCPM, além de participar ativamente de projetos de pesquisa em parceira

com outras universidades brasileiras.

O SCPM conta hoje com uma equipe de pesquisadores em tempo integral composta de 1

professor titular, 1 professor colaborador além de doutorandos, mestrandos, alunos de

iniciação científica e pessoal técnico de apoio. As atividades científicas desenvolvidas são

financiadas na sua maioria com recursos gerados através de projetos de pesquisa nacionais

e internacionais além da prestação de serviços e projetos em parceria com diversas

empresas. A estratégia de desenvolver seus projetos de pesquisa o mais próximo possível

das indústrias viabiliza uma rápida implementação dos resultados tecnológicos obtidos.

Reunir parceiros para desenvolver projetos mais arrojados tem sido a marca do trabalho do

SCPM, o que resultou em parcerias estratégicas desde a sua criação com o Institut für

Produktionsmanagement, Technologie und Werkzeugmaschinen (PTW) e com o Fachgebiet

Datenverarbeitung in der Konstruktion (DiK), ambos da Technische Universität Darmstadt na

Alemanha. Essas parcerias já resultaram em inúmeros projetos de pesquisa em conjunto e

um contínuo intercâmbio de alunos de graduação, mestrado e doutorado, além de

professores de ambos os lados.

Desde 2005 o SCPM possui também uma parceria com o Institut für Werkzeugmaschinen

und Fabrikbetrieb (IWF) da Technische Universität Berlin, Alemanha, e mais recentemente

com a Hochschule Rhein-Main em Rüsselsheim.

O SCPM dispõe de modernos recursos de hardware e software para o desenvolvimento dos

trabalhos de pesquisa atuando em quatro linhas de pesquisa: desenvolvimento integrado do

produto, usinagem com altíssima velocidade, monitoramento do processo de usinagem e

fábrica digital, além de oferecer suporte a pequenas e médias empresas para especificação,

escolha e implementação de sistemas CAD/CAPP/CAM/PDM.

Adicionalmente o SCPM possui uma Máquina de Medir por Coordenadas e um Sistema de

Calibração Laser Renishaw, que possibilitam o desenvolvimento de projetos de pesquisa

tanto com o foco na integração digital da cadeia CAD/CAM/CAQ, como também no

desenvolvimento de métodos para comparação da representação de superfícies complexas

nos sistemas CAD e o modelo real após a usinagem, permitindo a avaliação de estratégias

de corte e métodos de interpolação da trajetória da ferramenta.

Procurando atender às novas necessidades de empresas de pequeno e médio porte, o SCPM

iniciou trabalhos de pesquisa voltados ao Gerenciamento do Ciclo de Vida do Produto

(Product Data Management - PDM; Product Lifecycle Management - PLM). E hoje possui uma

V

plataforma de testes e uma equipe de alunos de graduação e pós-graduação (GAP - Grupo

de Aplicação de PDM) desenvolvendo simulações do processo de gerenciamento de dados

do produto ao longo de todo o ciclo de desenvolvimento.

Ainda dentro de seu objetivo de trabalhar com sistemas computacionais que representem o

estado da arte, o SCPM criou um grupo de trabalho para atuar no Planejamento Digital de

Processos tendo como foco o desenvolvimento de competências para atuar na temática

Fábrica Virtual e hoje já realiza projetos de pesquisa nesta área com renomadas empresas.

O material didático desenvolvido pela equipe do SCPM nas áreas de projeto e manufatura

auxiliados por computador tem sido utilizado não só nos cursos de engenharia da FEAU, mas

também por muitas outras universidades de diferentes lugares do Brasil. Esta atuação

pautada pelo trinômio pesquisa-ensino-extensão tem sido um importante processo re-

alimentador de todo o trabalho.

Desta maneira, o SCPM, além de uma forte inserção na área de pesquisa, tem conseguido

interagir de maneira positiva na definição das grades curriculares dos cursos de engenharia,

trazendo o que existe de mais inovador em desenvolvimento integrado do produto

contemplando desde a concepção até a manufatura.

Atualmente o SCPM desenvolve projetos financiados pelo CNPq e pelo BMBF-DLF e possui

outros em processo de avaliação junto às agências: CAPES, CNPq, FAPESP e DFG.

Mesmo enfrentando as dificuldades e os desafios inerentes à conjuntura brasileira e a uma

universidade particular, o projeto do SCPM visa uma formação ampla de seus pesquisadores

e estudantes, enfatizando o aspecto da pesquisa e a inserção internacional de sua equipe

através de intercâmbios, destacando-se assim dentro do projeto institucional como um

moderno provedor de serviços, dedicado às necessidades dos alunos que atuam no

laboratório, das indústrias com as quais tem desenvolvido projetos e da sociedade no seu

todo.



Lab. de Sistemas Computacionais para Projeto e Manufatura

Rod. Luis Ometto (SP 306), Km 24

13.451-900 Santa Bárbara d´Oeste, SP

Tel: (19) 3124-1792 Fax: (19) 3124-1788

E-mail: [email protected]

Home Page: http://www.unimep.br/scpm

VI

Parceiros SCPM

VII

Apresentação

O Laboratório de Sistemas Computacionais para Projeto e Manufatura (SCPM) realiza desde

1996 o Seminário Internacional de Alta Tecnologia, abordando temas focados em duas

grandes áreas: Manufatura e Desenvolvimento Integrado do Produto, alternadamente.

Dentro desses temas, seu Comitê Científico busca as inovações que estão sendo

implantadas com sucesso na indústria, e já no primeiro evento realizado trouxe para o Brasil

o tema da Usinagem com Altíssima Velocidade de Corte (High Speed Cutting - HSC).

Hoje este evento é reconhecido como um referencial no Brasil na divulgação de novas

tecnologias e métodos de trabalho, devido à atualidade e ao nível técnico dos temas

abordados, atraindo a atenção e a participação de pessoal técnico qualificado das mais

renomadas empresas localizadas no Brasil e de professores e pesquisadores de diversas

universidades.

Vencendo os desafios do desenvolvimento do Produto

Atualmente as empresas estão enfrentando grandes desafios devido à crescente

internacionalização da competição que vem sendo orquestrada por concorrentes mais

eficientes ou até mesmo mais agressivos, seja pela capacidade de inovação de alguns, seja

pelos baixos custos e escala de produção de outros. Isto tudo em um ambiente econômico e

político que demanda grande esforço e dedicação no planejamento estratégico visando o

futuro do país.

Para serem bem sucedidas as empresas precisam oferecer produtos inovadores, com time

to market reduzido, preços competitivos e com forte viés de sustentabilidade, características

estas que devem ser estabelecidas já na fase de desenvolvimento do produto e de forma

conectada com clientes e fornecedores.

Pesquisas sobre perspectivas nas áreas de Engenharia do Produto e Engenharia de

Processo demonstram que é necessário o acompanhamento minucioso, praticamente em

tempo real, do processo de desenvolvimento do produto, conjugado com o desenvolvimento

dos processos de produção com vistas a buscar a otimização do todo. Isso somente é

possível por meio da aplicação de tecnologias inovadoras nas várias fases do ciclo do

produto.

É considerando esse contexto que os organizadores do evento buscaram identificar as

inovações que vêm sendo geradas em projetos de pesquisa e desenvolvimento que já estão

sendo implantados com sucesso na indústria. Nos países desenvolvidos as perspectivas da

Engenharia do Produto apoiam-se em programas como o Industrie 4.0 na Alemanha e em

VIII

conceitos e tecnologias como Produtos e Sistemas Físico-cibernéticos, Smart Products,

aplicação de modelos humanos e o Frontloading, entre outros, no Japão e Estados Unidos.

É com essa visão que convidamos as comunidades industrial e acadêmica brasileiras para

participarem da 19ª edição do Seminário Internacional de Alta Tecnologia onde estaremos

discutindo os desafios que se apresentam no desenvolvimento do produto com o objetivo de

gerar novas ideias e soluções que serão determinantes para o sucesso das empresas.

Visando contribuir para a consolidação de um processo inovador no desenvolvimento de

produtos o evento deste ano abordará os seguintes temas:

4ª Revolução Industrial (“Industrie 4.0”)

Produtos e componentes físico-cibernéticos

Avaliação virtual da aplicação do produto

Modelos humanos biomecânicos

Sistemas PLM – ágeis e de arquitetura aberta

Frontloading e o desenvolvimento de novos produtos

Cooperação internacional universidade empresa

IX

Comitê Científico

Prof. Dr.-Ing. Klaus Schützer - SCPM - FEAU - UNIMEP, Brasil - Presidente

Prof. Dr. Alexandre T. Simon - PPGEP - UNIMEP - Editor Técn. Revista Máquinas e Metais

Prof. Dr. Alvaro J. Abackerli - IPT

Dr.-Ing. Bernd Pätzold - ProSTEP AG, Alemanha

Prof. Dr.-Ing. Dirk Bierman - IST - TU Dortmund, Alemanha

Prof. Dr.-Ing. Eberhard Abele - PTW - TU Darmstadt, Alemanha

Prof. Dr. h.c. Dr.-Ing. Eckart Uhlmann - Fraunhofer IPK, Alemanha

Prof. PhD. Elso Kuljanic - Università degli Studi di Udine, Itália

Prof. Tit. Dr.-Ing. Henrique Rozenfeld - NUMA - EESC - USP, Brasil

Prof. Dr. Jan Helge Bøhn - Virginia Tech University, Estados Unidos da América

Prof. Dr.-Ing. Michael Abramovici - Ruhr-Universität Bochum, Alemanha

Prof. MSc. Patrick G. Serraferro - Ecole Centrale de Lyon, França

Prof. Dr. Pedro Filipe Cunha - Escola Superior de Tecnologia de Setúbal, Portugal

Prof. Dr.-Ing. Rainer Stark - IWF - TU Berlin, Alemanha

Prof. Habil. Dr.-Ing. Ralph H. Stelzer - KTC - TU Dresden, Alemanha

Prof. Habil. Dr.-Ing. Reiner Anderl - DiK - TU Darmstadt, Alemanha

Prof. Dr. Reginaldo Teixeira Coelho - NUMA - EESC - USP, Brasil

Dr.-Ing. Teresa De Martino - European Commission - Directorate General XIII, Bélgica

Prof. Dr. Sc. Toma Udiljak - Croatian Association of Production Engineering, Croácia

X

Comitê Organizador

Prof. Dr.-Ing. Klaus Schützer – Presidente

Marcela Santana da Silva Romão – Secretária Executiva

Quinhones de Santana – Analista de Suporte

MSc. Carlos Eduardo Miralles

MSc. Renato Luis Garrido Monaro

Eng. Bruna Sakamoto

Eng. Dhiogenes dos Santos Sousa

Eng. Tiago Cacossi Picarelli

André Fagionatto de Castro

Felipe Alves de Oliveira Perroni

Florian Wolf

Gabriel Gaiotto Tezoto

Gabriel Abhener Medeiro Martins

Luiz Guilherme Luchiari Ferrari

Marcelo Octávio Tamborlin

Matheus Franco Soares

Sebastian Mack

Realização

Lab. de Sistemas Computacionais para Projeto e Manufatura Prof. Dr.-Ing. Klaus Schützer FEAU - UNIMEP

XI

Índice

Industrie 4.0 - Advanced Engineering of Smart Products and Smart Production ............................................................................................... 3

Prof. Dr.-Ing. Reiner Anderl, Technical University Darmstadt - DiK, Germany

Os Desafios e as Estratégias Nacionais para Tecnologia e Inovação ................................................................................................................ 29

Prof. Dr. Alvaro Toubes Prata, MCTI - Governo Federal, Brasil

Virtual Assessment of Product Use Based on Biomechanical Human Models ...................................................................................... 31

Prof. Dr.-Ing. Sandro Wartzack, KTmfk - FAU, Germany

Daniel Krüger, KTmfk - FAU, Germany

Jörg Miehling, KTmfk - FAU, Germany

Integrated Component Data Model for Smart Production Planning 59

Dipl.-Ing. André Picard, Technical University Darmstadt - DiK, Germany

Prof. Dr.-Ing. Reiner Anderl, Technical University Darmstadt - DiK, Germany

Cooperação com Universidades Alemãs: Oportunidades para a Indústria no Brasil ................................................................................ 81

Marcio Weichert, Centro Alemâo de Ciência e Inovação - DWIH, Brasil

Frontloading is a Key Success Factor and a Basis for Efficiency and Effectiveness of NPI Projects .............................................................. 95

Dr. Andreas Romberg, STAUFEN.Táktica - Consultoria.Academia Ltda., Brazil

Recent Approaches of CAD/CAE Product Development. Tools, Innovations, Collaborative Engineering ........................................... 119

Dr.-Ing. Peter Binde - Dr. Binde Ingenieure, Design & Engineering GmbH, Germany

XII

Additive Manufacturing in the Product Development ..................... 131

Dr. Jorge Vicente Lopes da Silva - CTI Renato Archer, Brazil

Bosch Engineering System – A Robust Design Process and 3D Model Applied in the Complete Product Development Chain ........ 173

MSc. Eng. Erwin Karl Franieck - Robert Bosch Ltda., Brazil

Processo de Desenvolvimento de Produtos Aeronáuticos ........... 191

MSc. Eng Waldir Gomes Gonçalves - Embraer S.A., Brasil

PLM na Magneti Marelli Cofap: Compartilhando um Caminho, Dificuldades e Desafios na Implantação Globalizada ..................... 203

Mauro Conceição - Magneti Marelli Cofap, Brasil

1

Artigos Técnicos

Technical Papers

2

Prof. Dr.-Ing. Reiner Anderl

Prof. Anderl was born in 1955 and studied mechanical engineering at

the Universität Karlsruhe, Germany, where he received his diploma in

1979. He received the Dr.-Ing. degree in Mechanical Engineering at

the Universität Karlsruhe in 1984. From 1984 to 1985 he served as

technical manager of a medium sized company. He then returned as

a senior engineer to the Institute for Applied Computer Science in

Mechanical Engineering (RPK) at the Universität Karlsruhe. In 1991

he has habilitated and in 1992 he has received the Venia Legendi,

which includes the authorization to teach CAD/CAM technology. In

April 1993 he accepted the call for the professorship for computer

integrated design (Fachgebiet Datenverarbeitung in der Konstruktion,

DiK) at the faculty Mechanical Engineering, Technische Universität

Darmstadt,Germany. At the Technische Universität Darmstadt, he

served as the dekan (dean) of the faculty in Mechanical Engineering

from 1999 until 2001, during which time the new bachelors and

masters program in mechanical engineering, mechatronics, and

computational engineering was defined and implemented; the first of

such in Germany. He has served on numerous faculty and university

committees and commissions, including serving as prodekan where

he managed the mechanical engineering faculty business office. He

is a member of the Zentrale Evaluierungs- und Akkreditierungs-

agentur (ZEvA), a national accreditation council based in Hannover,

Germany, where he works on issues related to bachelor- and master-

program accreditation. Prof. Anderl has served as vice president at

the Technische Universität Darmstadt from January 2005 until

December 2010.

[email protected]

DiK, TU Darmstadt

The Department of Computer Integrated

Design (DiK) is part of the Faculty of

Mechanical Engineering of the Technische

Universität Darmstadt. The integration of

information technology as integral part of

modern mechanical engineering and the

linkage of research and education to

industrial needs are our fundamental targets.

The principles and methods of processing

product data even today are developing

rapidly. To understand product data, product

data flows and product data processing, a

holistic approach named Product Data

Technology (PDT) has been chosen for

education and research. The scientific

strategy of the DiK is based on four main

research fields: “Information Modeling”,

“Virtual Product Creation", ”Collaborative

Engineering” and “Digital Factory. These

research fields contribute significantly to the

scientific progress of Virtual Product

Development and Virtual Factory and support

the creation of advanced competencies to

enable new innovation and strengthen

industrial competitiveness.

3

Industrie 4.0 - Advanced Engineering of Smart

Products and Smart Production

Abstract

Industrie 4.0 is a strategic approach for integrating advanced control systems with internet

technology enabling communication between people, products and complex systems. The

key approach is to equip future products and production systems with embedded systems

as a basis for smart sensor and smart actuators for enabling communication and intelligent

operation control. These so-called Cyber-Physical-Systems challenge design and

development processes and require appropriate engineering approaches. Within this

contribution the state–of-the-art for Industrie 4.0 is being presented, key use cases are

reported and an approach for establishing Industrie 4.0 in industry is presented. In this

context, a fundamental issue is to understand the role of integrated safety, security, privacy

and knowledge protection.

Keywords

Industrie 4.0; smart engineering; smart sensors; smart actuators; safety and security;

multidisciplinary product development; mechatronics; adaptronics, cyber-physical

systems.

19º Seminário Internacional de Alta Tecnologia Inovações Tecnológicas no Desenvolvimento do Produto

4

1 Introduction

Industrie 4.0 implies the 4th industrial revolution and is one of the German research initiatives

to implement the German high-tech strategy 2020 [1] to meet the challenges of the 21st

century. While the 1st industrial revolution is considered as the introduction of hydro power

and steam power, the 2nd industrial revolution is understood as the introduction of mass-

production techniques by using electric energy. The 3rd industrial revolution is based on the

application of electronic systems and information technology for enhancing manufacturing

automation. A significant breakthrough is now expected as the 4th industrial revolution by

introducing so-called cyber-physical systems (see Figure 1) [2, 3, 4, 5].

Figure 1: Industry 4.0 - The 4th industrial revolution (source:Zukunftsprojekt Industrie 4.0 [3])

The fundamental approach of Industrie 4.0 is using the ability of cyber-physical systems to

provide intelligence and communication for artificial, technical systems, which then are called

smart systems. Smart systems may be understood as a consequent successor technology of

mechatronic and adaptronic systems. The main feature is the integration of cyber-physical

systems for enabling inter-system communication and self-controlled system operation. Smart

systems are to be used for condition monitoring, structural health monitoring, remote diagnosis

and remote control. They are a kernel component for smart products, smart factories, smart

grids, smart logistics or even the smart city (see Figure 2). The intension for introducing smart

systems is the establishment of new value-added processes and new value-added networks

to increase and to improve flexibility, adaptability and efficiency of business processes.


5

Figure 2: The smart system approach

Business processes based on smart systems will also open the gate to establish

fundamentally new business models where the functionality of smart systems will be extended

with integrated services. This new packaging of systems’ functionality and services will enable

new approaches to meet customer and market demands.

Within this contribution the approach of Industrie 4.0 will be explained and use case scenarios

will be presented. Furthermore, the approach of transferring Industrie 4.0 to the industry in

Germany will be illustrated as well as some international activities to strengthen industrial

competitiveness based on smart systems.

2 Fundamental Approaches of Industrie 4.0 Technology

Industrie 4.0 technology aims at enabling communicating, intelligent and self-controlled

systems. From a technological point of view, Industrie 4.0 is characterized by 4 fundamental

conceptual approaches. They comprise:

Cyber-physical systems,

Internet technology,

Components as information carriers and

Holistic safety and security including privacy and knowledge protection.

The combination of these conceptual approaches enable smart systems as a kernel feature

of Industrie 4.0 applications.


6

2.1 Cyber-Physical Systems

The approach of cyber-physical systems has been described by LEE [6] as an intersection of

the theory of computation and the theory of dynamic systems. This results in two

complementary approaches called

Cyberizing the physical:

Cyberizing the physical aims at specifying physical subsystems with computational

abstractions and interfaces. This also leads to equip physical subsystems with

intelligence enabled e.g. through embedded systems. Furthermore, communication

becomes an important feature to interact with both, other cyber-physical systems as well

as humans.

Physicalizing the cyber:

Physicalizing the cyber expresses abstractions of dynamic systems to software and

interfaces as well as network components to represent their dynamic behavior in time.

Cyber-physical systems may be understood as a consequent configuration of embedded

systems, sensors, actuators including network access. Figure 3 shows a configuration

approach that enables the creation of cyber-physical systems and its further application as

cyber-physical production systems.

Figure 3: Cyber-physical systems

In cyber-physical systems network access can be provided in particular by equipping

embedded systems with an internet protocol address (IP address).


7

2.2 Internet Technology

Modern and future internet technology provides essential approaches to enhance the

performance of cyber-physical systems. These internet technology approaches comprise 3

concepts:

The internet of things (IoT):

The internet of things comprises communicating smart systems using IP addresses. The

upcoming IPv6 (internet protocol version 6) supports an IP address space of 128 bits

which enables to define 2^128 individual addresses or 3.4*10^38 addresses. This

enables each and every physical object being equipped with a unique IP-address.

The internet of services (IoS):

The internet of services comprises new service paradigms such as provided by the

service oriented architecture (SOA) [11] or the REST-technology [8] .

The internet of data (IoD):

In an environment of the previously mentioned internet of things and internet of services

technologies huge amount of data will be generated. The internet of data will enable to

transfer, to store mass data appropriately, and to provide new and innovative analysis

methods to interpret mass data in the context of the target application.

Figure 4 illustrates the internet technology impact on cyber-physical systems.

Figure 4: Impact of modern and future internet technology


8

As a consequence new industrial topics such as big data and could computing are gaining an

increasing importance.

2.3 Manufacturing Objects as Information Carriers

The approach of cyber-physical systems enables objects to be identified, localized and

addressed. Assigned to manufacturing objects such as single parts and assemblies this

technology opens new innovation paths. Manufacturing object become information carriers as

well as connected objects in a network of communicating instances. Manufacturing history

assigned to manufacturing objects create individual object information which is essential for

successive manufacturing processes and backtracing analysis.

Furthermore, manufacturing objects are connected to product model structures as well as

process planning data and thus they are enabled to actively control their own manufacturing

processes and procedures. Figure 5 illustrates an example of a manufacturing object, the

bottom of a pneumatic cylinder. The bottom part is identified and by analyzing its product

structure, its assembly is detected and through the assembly the assigned assembly plan, the

assembly area as well as the appropriate counter parts are accessed. This scenario also

shows how optimization for assembly processes is supported.

Figure 5: Manufacturing objects as information carriers

To support the concept of manufacturing objects becoming information carriers an appropriate

specification of information attached to manufacturing objects is required. This requirement

can be met by specifying a so-called component data model. The component data model is

derived from the product data model approach as available through the STEP standard

(Standard for the Exchange of Product Model Data, ISO 10303 “Product Data Representation

and Exchange [7]).


9

2.4 Holistic Approach for Safety, Security, Privacy and Knowledge Protection

Cyber-physical systems equipped with internet technology (IoT, IoS and IoD) require

outstanding concepts and technologies for to ensure safety, security, privacy and knowledge

protection. These concepts have to be applied in a real-time environment, which typically

addresses manufacturing environments. Figure 6 compares issues of the office-oriented

environments with those issues typically addressed by manufacturing environments.

Figure 6: Synchronous versus asynchronous application environments [9]

Safety requirements address the continuously available manufacturing operation ability while

security aims at the resilience against external and internal attacks against the cyber-physical

environment.

Privacy ensures the execution of operational functions without being monitored to a third

instance. Furthermore, knowledge protection provides methods and tools to avoid access to

manufacturing knowledge from outside or from non-authorized instances.

Clearly, Industrie 4.0 concepts require IT safety and security to be tied closely to physical

manufacturing processes also meeting real-time requirements.

3 Use Case Scenarios

Industrie 4.0 is expected to change the industry significantly. One of major changes is to

further develop process management, which today is strongly depending on centralized

methods to more de-centralized but interlinked methods. Planning and control of processes

will become much more flexible, adaptable and resilient against disruption. Some use case

scenarios illustrate expected benefits.


10

3.1 Use Case 1 “Component as an Information Carrier”

Provided that components as manufacturing objects are identifiable or even addressable their

current status will be registered individually. Use case 1 addresses manufacturing failures and

its effect analysis. After having manufactured the bottom of the pneumatic cylinder the part is

checked and verified against product design data in particular its dimensions and tolerances.

The measurement indicates dimensions not meeting tolerance requirements (see Figure 7).

Figure 7: Checking and verifying manufactured parts

As a consequence the reasons why tolerances are not met have to be detected. Therefore,

effect analysis is appropriate and in this case the manufacturing plan indicates which

manufacturing process is responsible for the failure. Through the manufacturing process the

machine tool and the operation tool are identified and their conditions are being analyzed (see

Figure 8).

In this use case it becomes evident that the operation tool (red curve in the right picture of

Figure 8) approaches the end of its lifetime.

Figure 8: Manufacturing plan and monitoring of the operation tool


11

3.2 Use Case 2 “Process and Condition Monitoring”

Process and condition monitoring is one of the most important use cases. As products or

components of products are based on cyber-physical systems their smart sensors are able to

deliver data about the products’ or the components’ condition. Such data might be e.g.

temperature, strain or vibration. The analysis and assessment of the data streams deliver

information about the products’ or components’ condition. Consequently, process monitoring

methods can be supplied with information indicating the process stability or instability [10].

Actions to ensure process stability such as load balancing and predictive maintenance can be

taken. Figure 9 illustrates process monitoring in a manufacturing environment.

Figure 9: Process monitoring in a manufacturing environment

While this use case has a high impact on the efficiency of respective value-added processes,

this use case also underlines the high importance of analyzing promptly data streams

produced by smart sensors. This is considered as an area where significant development of

new innovation is expected.

3.3 Use Case 3 “Additive Manufacturing”

A very promising use case comprises additive manufacturing. This upcoming technology uses

3D-CAD data to drive a layer-by-layer manufacturing process. Data representing a spare part

is delivered from a 3D-CAD system and is exported either as a STEP file or as a STL file. This

description of the parts’ geometry is used to compute ordered slices of the parts’ geometry.

For each of these slices the tool path is generated and used to control the tool operation to

produce the part layer by layer (see Figure 10).


12

Figure 10: Additive manufacturing based on fused deposition manufacturing

The attractiveness of these use case results from locating additive manufacturing centers in

the main markets worldwide and controlling the production by sending manufacturing control

data to the appropriate additive manufacturing unit. Equipped with cyber-physical systems the

additive manufacturing unit will report back about the successful production or in case of any

problems also reports will be sent back. This scenario is in particular of interest for producing

spare parts (Figure 11).

Figure 11: Use case “Additive Manufacturing”

Furthermore, also statistical data could be reported back indicating the frequency of produced

parts in the various market worldwide.


13

4 Transferring Industrie 4.0 to Industry

Industrie 4.0 an important research initiative to implement the German high-tech strategy. A

report about the transfer of the conceptual approach of Industrie 4.0 to German industry has

been delivered to the German government in spring 2013. Industry itself has taken action to

drive the implementation of Industrie 4.0. The main activity is the establishment of the so-

called Platform Industrie 4.0 under the organizational auspices of 3 industrial associations

BITKOM (ICT industry), VDMA (mechanical and process industry) and ZVEI (electrical and

automation industry). Figure 12 shows the organizational structure of the Platform Industrie

4.0.

Figure 12: Platform Industrie 4.0 [14]

A major approach is the collaboration between industry and the scientific community through

the initiation of a scientific advisory board. An important contribution of the scientific advisory

board was the definition of 17 theses explaining the main features of Industrie 4.0 (see Figure

13).

In the meantime, a research roadmap has been published and a number of research projects

have been initiated to contribute to the Industrie 4.0 technology. Furthermore, a couple of

demonstration centers for Industrie 4.0 have been established (Figure 14).

Industrie 4.0 has created awareness in both, German industry and academia. The technology

development, however, has also been started internationally. The European Commission has

published a research initiative on advanced manufacturing [12], in the United States the

Industrial Internet Consortium [13] has been established and further initiatives have been

started in China, South Korea and Japan. This confirms the strong movement to make the

Internet of Things, the Internet of Services and the Internet of Data a reality.


14

Figure 13: Theses of the scientific advisory board of the Platform Industrie 4.0 [15]

Figure 14: Research roadmap and demonstration centers for Industrie 4.0 [14]

5 Summary

Industrie 4.0 has become an important initiative for German industry. It is of strategic

importance and aims at strengthening industrial competitiveness. The main features driving

Industrie 4.0 are the development of cyber-physical systems, the integration of the internet of

things, the internet of services and the internet of data technology, the understanding of

components being information carriers and the implementation of a holistic approach to

ensure safety, security, privacy and knowledge protection.


15

The main target is to improve the value-added processes and to develop new business models

for strengthening industrial competitiveness. Future research activities will focus on smart

systems development, vertical and horizontal process integration and seamless digital

integration of lifecycle phases.

6 References

[1] N.N.

Hightech-Strategy 2020 for Germany

http://www.hightech-strategie.de/en/350.php

access: August 13th, 2012

[2] N.N.

Industry 4.0

http://www.hightech-strategie.de/de/2676.php

Access: August 13th, 2012

[3] Kagermann, H.; Wahlster, W.; Held, J.; (Hrsg.)

Bericht der Promotorengruppe Kommunikation. Im Fokus: Das Zukunftsprojekt

Industrie 4.0. Handlungsempfehlungen zur Umsetzung

Forschungsunion, 2012

[4] Broy, M.; (Hrsg.)

Cyber-Physical Systems. Innovation durch Software-intensive eingebettete

Systeme

acatech DISKUTIERT. Springer Verlag, Berlin Heidelberg, 2010

[5] acatech (Hrsg.)

Cyber-Physical Systems. Innovationsmotor für Mobilität, Gesundheit, Energie und

Produktion

acatech POSITION. Springer Verlag, Berlin Heidelberg, 2011

[6] Lee, E. A.: CPS Foundations. In: Proceedings of the 47th Design Automation

Conference (DAC). ACM/IEEE, June, 2010, S. 737 – 742

[7] Anderl, R., Strang, D., Picard, A., Christ, A.

Integriertes Bauteildatenmodell für Industrie 4.0 – Informationsträger für cyber-

physische Produktionssysteme. In: Zeitschrift für den Wirtschaftlichen

Fabrikbetrieb (ZWF), 109 (2014, 1-2), 64-69.

[8] Steinmetz C., Christ, A., Anderl R.

Data Management based on Internet Technology using RESTful Web Services In:

Proceedings 10th International Workshop on Integrated Design Engineering

IDE September 2014, Gommern, Germany


16

[9] Grimm, M.;., Anderl R.; Wang, Y.

Conceptual Approach for Multidisciplinary Cyber-Physical Systems Design and

Engineering. In: Proceedings of TMCE 2014, ISBN 978-94-6186-177-1

Budapest Hungary 2014

[10] Picard, A.; Anderl, R.

Smart Production Planning for Sustainable Production based on Federative

Factory Data Management. In: Proceedings of TMCE 2014 (Horvath I., Rusak Z.,

eds.). Budapest , pp. 1147-1156. ISBN 978-94-6186-177-

[11] Picard, A.; Anderl, R.; Schützer, K.; Moura, A. Alvaro de Assis :

Linked Product and Process Monitoring in Smart Factories based on Federative

Factory Data Management. ASME 2013 International Mechanical Engineering

Congress and Exposition, Volume 11: Emerging Technologies San Diego,

California, USA

[12] N.N.

Adavanced Manufacturing – Advancing Europe: Report of the Task Force on

Advanced Manufacturing and Clean Production

http://ec.europa.eu/enterprise/flipbook/ADMA/#/1/zoomed

European Union 2014; accessed August 15th, 2014

[13]N.N.

Industrial internet Consortium

http://www.iiconsortium.org/

Accessed August 15th, 2014

[14]N.N.

Plattform Industrie 4.0

http://www.plattform-i40.de/


[15]N.N.

acatech: Thesen des Wissenschaftlichen Beirats

http://www.acatech.de/fileadmin/user_upload/Baumstruktur_nach_Website/Acatec

h/root/de/AktuellesPresse/PresseinfosNews/ab_2014/Industrie_4.0_Broschuere.pd

f



17


18


19


20


21


22


23


24


25


26


27

28

Prof. Dr. Alvaro Toubes Prata

Secretário Executivo do Ministério da Ciência, Tecnologia e Inovação.

É professor titular do Departamento de Engenharia Mecânica da

Universidade Federal de Santa Catarina, UFSC. Possui graduação

em Engenharia Mecânica e em Engenharia Elétrica pela

Universidade de Brasília, mestrado em Engenharia Mecânica pela

Universidade Federal de Santa Catarina e doutorado em Engenharia

Mecânica pela Universidade de Minnesota, EUA. Há 36 anos na

UFSC, atua na graduação e pós-graduação, coordenando projetos de

ensino, pesquisa e extensão. Já publicou mais de 230 artigos

científicos completos em periódicos e anais de congressos, orientou

41 dissertações de mestrado e 20 teses de doutorado - possui duas

patentes depositadas. Em função de sua reconhecida atuação em

pesquisa e ensino em nível de pós-graduação, é pesquisador nível

1A no CNPq. De 2000 a 2004 foi pró-reitor de pesquisa e pós-

graduação da UFSC e ocupou a presidência do Fórum Nacional de

Pró-Reitores de Pesquisa e Pós-Graduação das Instituições de

Ensino Superior. É reconhecido com a Comenda da Ordem Nacional

do Mérito Científico - Classe Grã Cruz, dirigida a personalidades que

se distinguem por relevantes contribuições à ciência. Recebeu o

Prêmio Anísio Teixeira por ocasião do 60 aniversário da CAPES, em

reconhecimento à sua grande contribuição ao desenvolvimento das

Instituições Educacionais, Científicas e Tecnológicas no Brasil, por

meio do magistério, da pesquisa e da liderança institucional. De maio

de 2008 a maio de 2012 foi reitor da UFSC e por dois mandatos

ocupou a Vice-Presidencia da Associação Nacional dos Dirigentes

das Instituições Federais de Ensino Superior. É membro titular da

Academia Brasileira de Ciências, e coordena o Instituto Nacional de

Ciência e Tecnologia em Refrigeração e Termofísica. Suas áreas de

pesquisa são transferência de calor e mecânica dos fluidos.

[email protected]

MCTI – Governo Federal

O Ministério da Ciência, Tecnologia e

Inovação (MCTI) foi criado pelo Decreto

91.146, em 15 de março de 1985,

concretizando o compromisso do presidente

Tancredo Neves com a comunidade

científica nacional. Sua área de competência

está estabelecida no Decreto nº 5.886, de 6

de setembro de 2006. Como órgão da

administração direta, o MCTI tem como

competências os seguintes assuntos: política

nacional de pesquisa científica, tecnológica e

inovação; planejamento, coordenação,

supervisão e controle das atividades da

ciência e tecnologia; política de

desenvolvimento de informática e

automação; política nacional de

biossegurança; política espacial; política

nuclear e controle da exportação de bens e

serviços sensíveis.

29

Os Desafios e as Estratégias Nacionais para

Tecnologia e Inovação

30

Prof. Dr.-Ing. Sandro Wartzack

Prof. Wartzack, born in 1966, studied production engineering at the

Friedrich-Alexander-Universität Erlangen-Nürnberg, Germany, where

he received his diploma in 1994. From 1995 to 2000 he served as a

scientific assistant in engineering design at the same university. After

he had received his doctoral degree in 2000, he held several positions

at Brose Fahrzeugteile, a leading supplier in the automotive industry.

During that period he gained experience as a project engineer for

vehicle door systems. Later he became director of simulation and

knowledge management. Since 2009 he is head of the Department of

Engineering Design (Lehrstuhl für Konstruktionstechnik) at the FAU

Erlangen-Nürnberg. His research focuses on virtual and knowledge-

based engineering, user-centered design, lightweight design as well

as roller bearings and tribological coatings. Prof. Wartzack is member

of several national and international scientific committees including

e.g. Design Society and TechNet Alliance.

[email protected]

KTmfk - FAU

The mission statement of the Department of

Engineering Design (KTmfk) of the Friedrich-

Alexander-Universität Erlangen-Nürnberg is

“Design for Environment, Health and Safety”.

The KTmfk therefore works on the

development of optimal producible, resource

efficient and robust systems which address

the user’s needs in all research activities at

KTmfk. These activities include methods and

tools for optimized and shortened CAx

processes (CAD, CAE, KBE, DHM), for

dimensional management as well as user-

centered design. A further focus is on

practical examinations on the several test

facilities at the institute with emphasis on

roller bearings and tribological PVD-/PACVD-

coatings. The KTmfk is eagerly interested in

knowledge transfer between university and

industrial practice. Opportunities for

cooperation can be found in all fields of

research of the institute. KTmfk also offers a

variety of well-coordinated and modern

university courses for the design education of

students from engineering degree programs.

31

Virtual Assessment of Product Use Based on

Biomechanical Human Models

Abstract

For a long time the dominating view on product design was affected basically by functional

and economic aspects. Today a growing awareness of health in society is emphasizing

the importance to put the human into the center of all considerations. Indeed the basic

theme of a user-centered design is to provide an optimal fit between human beings and

technical systems by adjusting the properties of products to harmonize with the individual

competencies and needs of the users. In order to implement this idea, detailed information

on the prospective use of the product is needed already in the early stages of the

development process. This especially comprises the direct interaction between the user

and the product. Focusing on the assessment of ergonomic product properties like comfort

or safety, the question is to what extent the human body is stressed during the interaction

with a product. In this contribution simulations on the basis of biomechanical digital human

models are proposed to determine these physiological stress indicators. However since

the simulation procedure is based on methods and tools that were originally developed for

the purpose of medical motion analysis a meaningful application in product design requires

several issues to be addressed. Recent research therefore focuses on the integration of

biomechanical simulation models into common engineering environments. The objective

is to provide a straightforward way to describe and analyze the interaction between a

virtual product prototype and a virtual user model entirely within a CAD environment.

Therefore not only appropriate interfaces for the data exchange between multiple software

systems but also fundamental procedures to predict human motion based on interaction

goals are being developed and evaluated. Another crucial aspect is the adaptation of

human models to match the specific body characteristics and competencies of a certain

user group. Especially the latter aspect points out that the overall topic has an

interdisciplinary character. In this respect, the expertise of human scientists is

indispensable besides product designers and experts in numerical simulation.

Keywords

Ergonomics; virtual product development; design for use.

Authors

Prof. Dr.-Ing. Sandro Wartzack

Daniel Krüger

Jörg Miehling


32

1 Motivation

For a long time the dominating view on product design was affected basically by functional

and economic aspects. Today a growing awareness of health in society is emphasising the

importance to put the human into the center of all considerations. In fact the value of many

products is determined essentially by how well their properties harmonise with the individual

competencies and needs of the people who use them. [Dreyfuss 2003] The objective must be

to provide an optimal fit between human beings and technical systems. The idea of a user-

centred design is reflected in considerations that focus on aesthetics but most of all on product

ergonomics. This however requires detailed information on the prospective use, especially on

the direct interaction between users and products to be available already in the early stages

of the development process. The use of a product (Figure 1) is always associated with a task

the user wants to accomplish. Therefore a sequence of actions is chosen that trigger

appropriate functions of the product. At the same time the user constantly adjusts his

behaviour based on the perception of the products response. It is important to notice that

human behaviour (action) and product behaviour (response) are mutually dependent and

cannot be analysed separately.

Figure 1: User-product interaction

From an ergonomic perspective the matter of interest is how the organism of the user is

affected by the interaction process. Therefore the popular concept of load and stress [Bullinger

1994] is applied to product use. While interacting with a product a person can be subject to

external loads such as mechanical forces, vibrations, noise, chemical substances and also

cognitive loads (information). These external influences induce an internal stress on the

organism. Mechanical loads for example mainly affect the person’s musculoskeletal system

resulting in biomechanical stresses. The ergonomic load-stress concept is analogous to the

Environment

Interaction

product

technical

and

human-related

properties

user

demographic

and

psychographic

characteristicsactions

response

task


33

corresponding notions in engineering mechanics, which is shown by a simple example in

Figure 2. A cantilever beam is loaded by a force. The resulting deformation (stress) is a

function of the load but also of the physical properties Young´s modulus and moment of

resistance. Equally ergonomic stresses not only depend on the loads but also on the

biomechanical, physiological and psychological characteristics of the user.

Figure 2: Equivalence of the load - stress concept in engineering mechanics and ergonomics

Biomechanical stresses, e.g. muscular activity or joint reaction forces, can be associated with

ergonomic goals like comfort, safety and harmlessness. Hence if designers were able to

simulate the relationship between product characteristics and the level of stress prevalent

during the phase of use they would gain valuable insights on how to improve the ergonomic

quality of products.

In this paper therefore a novel application of biomechanical digital human models is proposed

to simulate ergonomic aspects of product use employing virtual product prototypes only. The

ob-jective is to provide a framework to analyse interaction processes between users and

products within a common computer-aided design (CAD) environment. Further it is shown

how user or user-group specific properties can be considered in biomechanical human

models. Even though the scope of this paper is on the design of products with close user

interaction like vehicle inte-rior, medical devices or sports equipment, many aspects are also

relevant for the design of workplaces and the planning of assembly processes.

2 Biomechanical Digital Human Models for Ergonomic Evaluation

In the early stages of the product development process industry standards like DIN 33401 can

provide general guidelines on issues of human-product interaction. However due to the

heterogeneity of human characteristics and the huge amount of imaginable products universal

rules on ergonomic design are often too generic as being helpful in a specific case. The

standard ISO 9241-210 therefore points out the importance of testing. Thereby a fully or partial

functional physical mock-up of the product is presented to persons in order to evaluate its


34

ergonomic quality. Even though the informational value of physical experiments is undoubted

they are always time consuming and costly. Consequently as in other areas of engineering

there is a strong motivation to replace physical testing with computer simulations. This gives

rise to digital human modelling. The idea is to have a virtual model of the user interacting with

a virtual prototype of the product. This not only leads to a reduction of costs but also gives

much more freedom to designers to think through multiple concepts because the results of a

virtual experiment are usually available within a short time span. Biomechanical simulation

systems like OPENSIM [Delp 2007] or ANYBODY [Damsgaard 2006] were developed to describe

structure and motion of the human musculoskeletal system based on multibody dynamics.

The skeleton is modelled as a set of rigid or partially compliant bodies that are interconnected

by joints. Muscles, tendons and ligaments are represented by special force actuators. Some

advanced muscle models even consider effects of fatigue. The primary purpose of these

software packages is the analysis of human motion sequences employing inverse dynamic

calculations (Figure 3). This method requires that the motion of the human model is

unambiguously determined by time series of the generalised coordinates and their derivatives

which correspond to the angles of human joints. If further all external forces F acting on

the body are known the equations of motion can be solved for the actuating forces T which

are identified as the joint torques generated by the muscles. In subsequent postprocessing

steps additional indicators of biomechanical stress like the level of muscular activity,

metabolism and joint reaction forces can be determined. Even though biomechanical digital

human models were developed for applications in motion medicine, they perfectly fit into the

ergonomic load-stress concept. Since they reveal the physiological causes for ergonomic

issues, the simulation results (e.g. muscular activity) can directly be regarded as ergonomic

assessment criteria.

Figure 3: Inverse dynamic analysis of motion

Being among the first to suggest biomechanical human models as design tools [Rasmussen

2003] used the ANYBODY modelling system to optimise the ergonomic properties of a hand

saw. Also the US Defense Advanced Research Projects Agency supports an effort to use

OPENSIM for de-sign activities within the scope of developing special suits for soldiers that


35

reduce the risk of injuries and fatigue in combat missions [Simtk 2013]. A broader application

in design however may be inhibited by the fact that biomechanical simulation systems have

not been integrated into the processes of virtual product development. Inverse dynamic

simulations determine the bio-mechanical stresses as a consequence of posture, motion and

external forces. The problem is that in case of a truly virtual simulation the information on how

a user will move during the interaction with a product is not available. A possible solution is to

record the motion of a test per-son and map the data on the human model [Robert 2013] but

this would again mean that a physical experiment had to be conducted and the major benefits

of virtual testing would be lost. In this paper therefore a concept for a CAD integrated

biomechanics laboratory is presented. The objective is to improve the usability of

biomechanical simulations in design. Usability in this context means that the person that uses

a simulation program must be able to provide the input data required to setup the

computations as well as understand and interpret the results. Design engineers usually know

very precisely what functions of the product have to be triggered in or-der to fulfil the tasks it

has been designed for. A formulation of the interaction processes be-tween the user and the

product should therefore rely on task descriptions that encode infor-mation on how the state

of the product or the environment has to be manipulated. Human ac-tions (posture and motion)

that achieve these manipulations should not be required as input in-formation but be predicted

by the simulation. Since a huge proportion of the synthesis work in design is nowadays done

using computer-aided methods it is reasonable to postulate a close integration of

biomechanical simulations with CAD engineering environments. In this regard user or user

group-specific properties are to be taken into consideration in the preceding digital hu-man

modelling step, eventually facilitating robust designs concerning ergonomic aspects of the end

product.

3 Consideration of User-Specific Characteristics in Biomechanical

Digital Human Models

As biomechanical digital human models, also called musculoskeletal models, comprise just a

skeleton as well as muscles, in the first stage we focus on differences in anthropometry as

well as motor functions like strength, range of motion and motion speed. [Miehling 2013]

proposed a process for the conception of biomechanical digital human models taking the

interdependences of the considered domains into account. This procedure is outlined in figure

4. The relevant data to consider the heterogeneous differences over the human life span in

the conception stage are taken from literature.

First of all gender and age of the model to be generated have to be chosen according to the

target group of the product to be developed. In the consecutive steps of the adaption process

percentile values in conjunction with already existing population data are favourable to specify

the model. However, data from manual measurements can be beneficial to model a specific

person or user group.


36

The method of choice for specifying the body measures in most cases is to chose a

percentile. The height and consequently scaling information for the body segments can then

be computed using population data taking into account the specified age and gender. Body

height data of one culture and gender in most cases can be presumed to follow a normal

distribution. Studies therefore mostly make the body height distribution of a specific age group

and sex available as mean value accompanied by the standard deviation or standard error as

for example in the National Nutrition Survey II (NVS II) published by the Max Rubner-Institut

in 2008. This representative survey among others reports the sociodemographic

characteristics as well as anthropometric measures of the german population. In contrast, if

the model to be generated should resemble a specific (real) person, manual measurements

have to be conducted to get the overall body height as well as values for the dimensions of

the individual body segments. Regardless of the data origin, subsequently scaling factors for

all body segments are calculated which are eventually used to scale the musculosceletal

model.

Figure 4: Overview of the conception process [Miehling 2013]

Another sophisticated method is to retrieve the segmental lengths through optical, marker-

based or markerless measurement systems usually used for motion capture purposes.

[Krüger 2012] for example developed a system for the markerless capture of motions as well

as scaling of biomechanical digital human models using the Microsoft Kinect sensor, originally

developed as game controller. This system automatically provides the scaling factors for the

body parts without need for further computation. Its user interface is depicted in Figure 5.

After collecting the data for the segmental lengths, the body weight of the model to be

generated has to be specified. Even though body weight is usually not normally distributed,

the majority of surveys report the weight in the same way as the body height and therefore

neglect valuable information about the underlying sample. Moreover, body weight tends to

increase with body size. This correlation hampers the computation of the body weight using

the body weight distribution of the population.

age / gender

body height /segment lengths

body weight /segment weights

range of motion strength

mass momentsof inertia

motion speed


37

Figure 5: Low-cost motion capture system based on Microsoft Kinect

In the present approach body weight can therefore either be chosen directly or by the body

mass index (BMI). [Keys 1972] advised the BMI (body weight [kg] / (body height)² [m²]) as a

measure for the physical constitution of populations. It removes the dependency between

weight and height and is a good predictor for body fat percentage. [Hemmelmann 2010]

calculated the BMI distribution for both genders and every single year of age using the LMS

method based on the raw data of the NVS II. After chosing a percentile of the BMI distribution,

the BMI can be computed. Hereafter, the body weight can be calculated taking into account

the body height of the preceding step. The entire body’s mass distribution, respectively

individual body segment weights, are then computed considering the scaling factors for the

body part dimensions. If the human model’s dimensions are scaled just considering the overall

change in body height, the mass distribution stays unaffected. The mass moments of inertia

of the individual body segments are especially important in dynamic simulations. If the

changes in the segments’ mass and dimensions are known, the inertia tensors can be

calculated. The maximum isometric forces generated by skeletal muscles are highly

dependent on age, weight and size. A taller, heavier person tends to be able of generating

bigger muscle forces in comparison to a shorter, lighter person of the same age, gender and

ethnicity. From around 30 years of age on, the maximum muscle forces decrease steadily.

Women are generally less strong than their male counterparts. [Stoll 2002] measured and

percentiliced the maximum isometric voluntary joint torques for a healthy population in a given

joint angle constellation. This data is used to scale the maximum isometric force of every

muscle of our biomechanical digital human model. Unlike with body measures, weight and

strength, there seems to be no clear correlation between the range of motion and the age of

a person. The distributions in this respect coincide largely, given that diseases like arthritis are

ignored. Due to the just stated aspects the range of motion is scaled using percentile values

without regarding the affiliation to a specific age group [Greil 2008]. The maximum motion


38

speed does not directly depend on body weight and size. The execution of movements

decelerates just a small proportion due to physiological changes in the skeletal muscles, but

largely due to the smaller maximum forces resulting from the progressing muscular dystrophy

with age. Additionally as weight increases, the segments’ mass moments of inertia rise and

therefore the same muscle forces yield lower angular accelerations and in turn angular

velocities. Taking the age distribution of the targeted culture into account, representative user

groups can be generated for the following virtual assessment of product use.

4 A CAD Integrated Biomechanics Laboratory

4.1 Concept and Related Work

In section 2 a task oriented formulation of user-product interactions and a seamless

integration with CAD environments have been identified as the most important

requirements on a virtual biomechanics laboratory. Existing approaches to connecting

biomechanical models and CAD systems are mainly addressing the problem of data

exchange. The Anybody modelling system e.g. provides an import filter for product geometry

created in SolidWorks. A closer integration can be achieved by coupling a complex

biomechanical model to the kinematics of a CAD integrated anthropometric model as

published by [Jung 2013] and previously by [Krüger 2012].

Figure 6: Concept of a CAD integrated biomechanics laboratory

Our concept (Figure 6) uses a similar idea: an anthropometric human model (skin model)

inside a common CAD environment serves as a front-end the design engineer uses for


39

preprocessing. Preprocessing mainly comprises the definition of geometric relationships

between the user and the product that are needed to setup the actual simulation performed

on the musculoskeletal model.

The concept is built on four important pillars: the transfer of anthropometric data and posture

(spatial registration) between the skin model and the musculoskeletal model, a task oriented

pro-tocol to formulate user-product interactions, a methodology to predict human motion and

the simulation of product behaviour. In the following sections these topics are illustrated by

means of a simple case study. The interaction process to be analysed is the situation of a

person driving a passenger car as depicted in Figure 7. Even though in reality the driver is

interacting also with the gear lever and the pedal, the only task considered in this example is

to turn the steering wheel slightly to the right. The question to be answered by this analysis is

which region of the body shows the highest muscular activity.

Figure 7: Case study: steering a passenger car

4.2 Transfer of Anthropometric Data and Posture

In the beginning of a virtual experiment all relevant characteristics of the user like body

measures or strengths must be choosen according to the procedure described in section 3.

The output of this process is a musculoskeletal model for the OPENSIM platform. For the case

study the musculoskeletal model has been choosen to resemble the anthropometric

properties of an average European male. From within the CAD environment this

musculoskeletal model is loaded and connected to the skin model. The connection consists

of a scaling operation and a subse-quent spatial registration of the two models. In the scaling

operation the limb lengths of the mus-culoskeletal model are determined by calculating the

distances of marker points located in the centres of the joints. These values are assigned to

corresponding CAD parameters of the skin model. Since the skin model is used as a front-


40

end of the actual simulation model it has to be assured that both models coincide

geometrically. This means that for example the location of a point defined on the skin model

must be unambiguously found also on the musculoskeletal model. This is achieved by a point

by point registration (see Figure 8): a set of datum points dis-tributed over the limbs of the

musculoskeletal model is fitted to a corresponding set located on the skin model by numerical

inverse kinematics. As a result the musculoskeletal model follows the posture specified by the

skin model.

Figure 8: Transfer of data between skin model and musculoskeletal model

4.3 Task Oriented Interaction Protocol

Once the human model is set up the next step is the formulation of the task to be analysed.

As postulated in section 2 this formulation should not rely on descriptions of human behaviour

(e.g. “move the right arm”) but on required manipulations of the product. Therefore a formal

interaction protocol has to be elaborated that could rely on a structured model of user-product

interaction as published by [Mieczakowski 2010]. The current prototype of the protocol

contains action goals and boundary conditions. To define an action goal the designer identifies

parts of the product model as human-machine interfaces and describes how these parts have

to be manipulated in order to trigger the desired function of the product. A boundary condition

is a geometric relationship between the user model and the product model or the environment.

In case of the steering example (Figure 9) the claims that the buttocks of the user remain on

the seat while both hands remain on the whee are typical boundary conditions whereas the

required rotation of the wheel link is an action goal.

The boundary conditions and action goals are the input parameters for an algorithm (motion

predictor) that predicts the motion the user will most likely choose to fulfil the task. This

algorithm is described in the next section.


41

Figure 9: Interaction protocol: boundary conditions and action goals

4.4 Prediction of Human Motion Based on the Optimality Principle

Optimality as the property of a system to maximise or minimise some function under given

con-straints is often found in nature. Examples are the minimisation of potential energy as a

driving force for chemical and physical processes or the assumption of natural selection

according to which form and behaviour of creatures are developing towards optima. Many of

the characteris-tics of human motor behaviour can be explained by the optimality principle. It

seems natural that humans perform movements in a way that minimum mechanical effort is

necessary. But also the elimination of kinematic and dynamic redundancy can be achieved if

those joint con-stellations and patterns of muscular excitation are preferred that entail less

effort compared to alternative solutions. Computer simulations based on the optimisation of

mathematical functions are therefore very promising approaches to predict posture and

motion of biomechanical human models.

The dynamics of the human musculoskeletal system can be described by equation (1).

(1)

Here (t) is the skeletons physical state vector consisting of the joint angles, the corresponding

generalised speeds as well as additional states of the muscles like tendon length, contraction

velocity and level of activation. The time derivative of the state depends on the current

state and the current control vector . Controls are time dependent neural signals that

activate the muscles and lead to torque generation in the joints. Under a simplified point of

view without considering muscles one can also directly apply torques to the joints and regard

these as controls. The evolvement of the state (= motion) is simulated by integrating equation

(1) over time. Hence predicting human motion means to determine a set of control signals


42

that lead to the achievement of a task defined by the interaction protocol described in section

3. Due to the kinematic and dynamic redundancies of the human musculoskeletal system

there is usually no unique solution to this problem. Instead one has to settle for finding a best

solution that minimises an objective function given in (2).

with (2)

This objective function assesses the motion (state and control) of the human model by means

of arbitrary optimality criteria encoded by the cost value that is associated with each time

step. Possible optimality criteria are discussed further below. The optimal control

signals are consequently the solutions of the following dynamic optimisation problem.

(3)

Optimal control problems of this type can be solved by several numerical methods [Todorov

2006]. A limiting factor for the application on complex dynamic systems like the human body

however is that most of the algorithms are computationally extreme costly.

An exception to this is the iterative linear quadratic regulator (iLQR) method that was originally

published by [Todorov 2005]. It would go beyond the scope of this contribution to cover all the

mathematical details. Instead only the coarse working principle of iLQR and how it is employed

to predict human motion within virtual testing of use is explained. The algorithm takes

advantage of the fact that for linear system dynamics and objective functions quadratic in u

the solution to the optimal control problem is relatively straightforward. Unfortunately

musculoskeletal dynamics are highly non-linear. The idea of iLQR is to iteratively use linear

approximations of the system dynamics function (1) and quadratic approximations of the

objective function (2) to construct a sequence of solutions that finally converges to the exact

solution. The methodology actually yields an optimal feedback controller which means that

not only the controls are determined but also feedback gains that could be used to correct

the motion from external disturbances. An iLQR controller was implemented on top of the

biomechanical simulator OPENSIM and applied to the case study introduced in section 4.1.

The task (turning the steering wheel) has been formulated using the interaction protocol

described in the previous section. The boundary conditions (buttocks on seat, both hands on

the wheel, feet on the pedals) are implemented by inserting five kinematic constraints into the

musculoskeletal model. The action goal (rotate the wheel to ) is used to derive the

optimality criteria for the iLQR controller. The resulting objective function is given in equation

(4).

+ (4)

The term in front of the integral only depends on the final time step T and penalises the

deviation of the gear levers actual position from the required position. In addition since the

motion should stop at the required position the velocity of the lever is required to be zero. The


43

integral term is called the running cost function since it assesses the way towards the target

position penalising the control effort for the muscles.

4.5 Simulation of the Product Behavior

User-product interaction processes as depicted in figure 1 are actually feedback loops. The

be-haviour of the product is affected by the behaviour of the user and vice versa. A simulation

for virtual testing of use must consequently also contain a behavioural model of the product.

Our concept permits a computational separation of the musculoskeletal user model and the

product model. Hence user behaviour and product behaviour can be processed in separate

simulation programs that however need to be synchronised to exchange data. The advantage

of this co-simulation approach over monolithic solutions [Damsgaard 2006] that handle the

product as a part of the multibody system employed to describe the user is that the behaviour

of the product is not limited to what can be described by multibody dynamics. In fact the

product simulator can be any kind of algorithm capable of emulating mechanical responses.

A problem of co-simulations is to define the system boundary of each simulation model. In

case of user-product interaction it is reasonable to define this boundary along the human-

machine interfaces of the product. In the case study the steering wheel can be identified as a

part of the human-machine interface. The product behaviour however is mainly determined

by the internal design of the steering system. This behaviour is sensed in terms of a reaction

force on the wheel. Hence the system boundary is the interface between the wheel and the

steering gear. In other words the wheel is treated as a component of the biomechanical

multibody tree while the reaction force generated by the steering system could be emulated

by an external product simulator. In this case the communication between the simulators

would be an exchange of wheel rotation and reaction torque. This approach entails the

necessity to export parts of the product model (the human machine interfaces) from the CAD

environment into the biomechanical simulation sys-tem. A prototypical interface between

OPENSIM and the CAD system CREO/PARAMETRIC (PTC) has been developed that allows

exporting arbitrary parts or sub-assemblies of the product model into the multibody simulator.

Therefore the mass properties of the parts and the kinematic con-straints to the surrounding

assembly are analysed automatically

5 Results of the Case Study

The CAD integrated biomechanics laboratory has been used to create a dynamic simulation

model for the car driving example introduced in section 4.1. The resulting motion sequence is

illustrated in Figure 10. Moreover the total sum over the control signals of the muscles that

are actuating the right shoulder joint is shown. This muscle group has been identified to

contribute the largest part of the actuation effort in the upper extremities of the body. Since

control signals are direct proportional to the level of muscular activitiy, they are ideal criteria


44

to assess the risk of muscular fatigue, which is one aspect of discomfort, during the user-

product interaction pro-cess. In the present case adjustments of the product design that lead

to a lower control values can therefore be regarded to improve the ergonomic quality of the

vehicle cockpit. However since the complete dynamic state trajectory is known, additional

stress indicators (e.g. joint loads) can be extracted in subsequent computations.

Figure 10: Case study: motion sequence and major actuation

6 Summary and Outlook

A growing awareness of health in society emphasises the importance of a user-centred design

process.

More than in former times design engineers will have to focus on product ergonomics. Since

ergonomic product properties are related to the interaction processes with the user, the im-

portance of testing for use is also growing. However traditional testing concepts are time con-

suming and costly because they usually require the manufacturing of physical mock-ups and

the conduction of experiments involving multiple test persons to cover the characteristics of

the tar-get user group. In this paper therefore biomechanical human models were proposed

as a possibility to simulate ergonomic aspects of user-product interaction already in the early

stages of the development process.

Hereby designers are enabled to predict and quantify the relationship between design

parameters and the level of biomechanical stress effects prevalent during product use within

the users organism. To improve the ergonomic quality of products the design is adjusted so

that stress indi-cators like muscular activity are kept at a moderate level. However the

application of biome-chanical simulations in design is currently not very widespread. The


45

dependence on experi-mental data for the specification of human behaviour and the

unsatisfying integration with exist-ing methods and tools of virtual product development were

identified as the main hurdles. The benefit of the concept for a virtual biomechanics laboratory

presented in this paper is the seam-less integration into an existing CAD/CAE environment.

Designers are not confronted with ex-perimental data and anatomical details on

biomechanical modelling. Due to the computational separation of product model and user

model it is possible to take advantage of a huge number of sophisticated CAE algorithms to

resemble the behaviour of the product. This is especially im-portant since many products

today are mechatronic systems that can’t be analysed using solely multibody dynamics. The

most crucial but also the most challenging aspect of virtual simulation of use is however the

prediction of human behaviour. Based on a task oriented formulation of user-product

interaction an optimal control algorithm is employed to synthesise the motion of the user. Even

though this is regarded a promising approach its validity has not been verified yet. Future

research will therefore have to address the experimental validation of motion prediction

methods. Equally the implementation of the concept presented is still incomplete. In particular

the task oriented interaction protocol and the computational interfaces to perform a co-

simulation of user model and product model require additional effort to become usable in

indus-trial applications.

Another important question is how designers have to interpret the results of a biomechanical

analysis. Biomechanical stress indicators at first glance tell little about what design changes

could improve the ergonomic properties of the product. The simulation system therefore

should provide the designer with guidance to design improvements by mapping the results

back into the space of design parameters.

7 References

Bullinger, H.; Ilg, R.: “Ergonomie: Produkt und Arbeitsplatzgestaltung“, Teubner Stuttgart,

Germany, 1994.

Damsgaard, M. et al.: “Analysis of musculoskeletal systems in the AnyBody Modeling

System”, Simulation Modelling Practice and Theory, Vol. 14, No. 8, 2006, pp

1100–1111.

Delp, S. et al.: “OpenSim: Open-Source Software to Create and Analyze Dynamic

Simulations of Movement”, IEE Transactions on Biomedical Engineering, Vol. 54,

No. 11, 2007, pp 1940-1950.

Dreyfuss, H.: “Designing for people: [the classic of industrial design]”, Allworth New York,

USA, 2003.


46

Greil, H.; Voigt, A.; Scheffler, C.: „Optimierung der ergonomischen Eigenschaften von

Produkten für ältere Arbeitnehmerinnen und Arbeitnehmer – Anthropometrie“,

Bundesanstalt für Arbeitsschutz und Arbeitsmedizin, Dortmund, 2008.

Hemmelmann, C.; Brose, S.; Vens, M.; Hebebrand, J.; Ziegler, A.: “Percentiles of body mass

index of 18-80-year-old German adults based on data from the Second National

Nutrition Survey”, Deutsche medizinische Wochenschrift (1946), Vol. 135, No. 17,

2010, pp 848-852.

Jung, M.; Damsgaard, M.; Andersen, M.; Rasmussen, J.: “Integrating Biomechanical

Manikins into a CAD Environment”, 2nd International Digital Human Modeling

Symposium, Ann Arbor Michigan, 2013.

Keys, A.; Fidanza, F.; Karvonen, M. J.; Kimura, N.; Taylor, H. L.: “Indices of relative weight

and obesity”, Journal of chronic diseases, Vol. 25, No. 6, 1972, pp 329-343.

Krüger, D.; Miehling, J.; Wartzack, S.: “A simplified approach towards integrating

biomechanical simulations into engineering environments”, 9th Norddesign

Conference, Aalborg, 2012, pp 334–341.

Max Rubner-Institut (Hrsg.): Nationale Verzehrsstudie II. Ergebnisbericht, Teil 1. Die

bundesweite Befragung zur Ernährung von Jugendlichen und Erwachsenen,

Berlin, 2008.

Mieczakowski, A.; Langdon, P.; Bracewell, R.; Clarkson, P.: “Toward a model of product-user

interaction: a new data modelling approach for designers”, International Design

Conference DESIGN 2010, Dubrovnik, 2010, pp 875–884.

Miehling, J.; Geißler, B.; Wartzack, S.: “Towards Biomechanical Digital Human Modeling of

Elderly People for Simulations in Virtual Product Development”, Human Factors

and Ergonomics Society 2013, International Annual Meeting, San Diego, 2013, pp

813-817.

Rasmussen, J. et al.:”Musculoskeletal modeling as an ergonomic design method”,

International Ergonomics Association XVth Triennial Conference, Seoul, 2003.

Robert, T.; Causse, J.; Denninger, L.; Wang, X.: “A 3-D dynamics analysis of driver’s

Ingress-Egress Motion”, 2nd International Digital Human Modeling Symposium,

Ann Arbor Michigan, 2013.

Simtk, “Warrior web project website”, https://simtk.org/home/opensim_ww, accessed

20.11.2013

Stoll, T.; Huber, E.; Seifert, B.; Stucki, G.; Michel, B. A.: “Isometric Muscle Strength

Measurement”, Thieme, Stuttgart, 2002.


47

Todorov, E.: “Optimal Control Theory”,in Doya, K.(ed.), Bayesian Brain, MIT Press

Cambridge, USA, 2006.

Todorov, E.; Weiwei, L.: “A generalized iterative LQG method for locally-optimal feedback

control of constrained nonlinear systems”, American Control Conference 2005,

Portland, 2005.


48


49

Nam

e 05

.32.

001

© LEHRSTUHL FÜR KONSTRUKTIONSTECHNIKFriedrich-Alexander-Universität Erlangen-NürnbergProf. Dr.-Ing. Sandro Wartzack 5

Chair of Engineering DesignFriedrich-Alexander-Universität Erlangen-Nürnberg

Sto

ckin

ger

HeadquartersSouthern Campus

SubsidiaryRöthelheim Campus

Executive BoardSchlossplatz

Faculty of EngineeringSouthern Campus


50


51


52


53


54


55


56


57

58

Dipl.-Ing. André Picard

Dipl-Ing. André Picard graduated from Technische Universität

Darmstadt in 2010. With the completion of his studies as Diplom-

Ingenieur, he joined in October 2010 his position as Research

Assistant at the Department of Computer Integrated Design (DiK) at

Technische Universität Darmstadt. Since midth of 2012 Mr. Picard is

part of the research field “Digital Factory” at the DiK. He was a member

of the DFG promoted project “Federative Factory Data Management”.

In particular he is engaged in the linking of different factory data and

the integration of mobile internet devices into the federative factory

data management. Since 2014 he takes part in the LOEWE centre

AdRIA (Adaptronic - Research, Innovation, Application). His research

covers methods for virtual development of adaptronic and cyber-

physical systems.

[email protected]

DiK, TU Darmstadt

The Department of Computer Integrated

Design (DiK) is part of the Faculty of

Mechanical Engineering of the Technische

Universität Darmstadt. The integration of

information technology as integral part of

modern mechanical engineering and the

linkage of research and education to

industrial needs are our fundamental targets.

The principles and methods of processing

product data even today are developing

rapidly. To understand product data, product

data flows and product data processing, a

holistic approach named Product Data

Technology (PDT) has been chosen for

education and research. The scientific

strategy of the DiK is based on four main

research fields: “Information Modeling”,

“Virtual Product Creation", ”Collaborative

Engineering” and “Digital Factory. These

research fields contribute significantly to the

scientific progress of Virtual Product

Development and Virtual Factory and support

the creation of advanced competencies to

enable new innovation and strengthen

industrial competitiveness.

59

Integrated Component Data Model for Smart

Production Planning

Abstract

The integrated component data model describes individual components throughout the

whole component life cycle from product development to recycling or disposal. Contrary

to the integrated product data model in which a generic product data model is instantiated

to multiple virtual product objects the integrated component data model focuses on a single

physical instance of real parts and assemblies generically called components. It therefore

combines individual data about the physical component with virtual product lifecycle

information.

Consequently the integrated component data model is a key-enabler for smart production

planning. Smart production planning is characterized by a bidirectional flow of information

between virtual product data on the one hand and physical component and environment

information of existing product generations on the other hand. In this context individual

component data actively influences preceding product development and production

planning processes.

In the context of the smart and resource-efficient factory of the future one main goal of

component-driven manufacturing is the improvement of resource exploitation. Aggregated

information provided in a schematic representation within the integrated component data

model facilitates the component-specific adaption of processes.

Within this paper the integrated component data model as well as an exemplary use case

for improved resource exploitation in the smart factory is presented.

Keywords

Integrated component data model; Industrie 4.0; computer aided manufacturing, cyber-

physical production systems.

Authors

Dipl.-Ing. André Picard

DiK, TU Darmstadt [email protected]

Prof. Dr.-Ing. Reiner Anderl

DiK, TU Darmstadt [email protected]


60

1 Introduction

In the past years manufacturing companies are facing a dramatic change. Individual customer

demands on highly customer-tailored goods and services are getting most important on the

global markets. Consequently the product variety increases while meantime the lot size

decreases. For such goods short time to market with innovative technologies at remarkable

low prices are crucial to successfully compete on the global markets.

The application of recent information and internet communication technologies to mechatronic

products in combination with today’s ubiquitous computing seems a promising solution for

these demands. So-called cyber-physical systems (CPS) are developed. They are digitally

enabled products which communicate and collaborate within ad-hoc networks of other cyber-

physical systems. They get the ability to autonomously decide and adaptively react on internal

and external events in order to optimize, diagnose and calibrate themselves and their

surrounding environment [1].

Digital product memories within those cyber-physical systems enable the seamless gathering

and use of component-individual on-line information throughout the whole life cycle from the

cyber-physical systems’ creation to recycling or disposal [2].

The provision of this information to engineers of the production planning is a key-enabler for

the creation of process knowledge. Engineers get the ability to extract interdependencies

between production planning and manufacturing processes at the shop floor first and to

consider this knowledge during production planning secondly. Therefor the continuous bi-

directional information flow throughout the whole component life cycle from product idea to

component recycling or disposal is needed.

To support such an information flow, an approach for the specification of the integrated

component data model and its integration in the engineering processes of the production

planning is presented within this paper. The integrated component data model enables front-

loading of production planning with derived knowledge from products, processes, resources

and component data. An exemplary use case on this smart production planning further

illustrates the benefits of the given approach.

2 Cyber-Physical Production Systems

Cyber-physical systems are an integration of mechanical, electronic components with recent,

internet based information and communication technologies [3]. Lee describes two

complementary approaches for cyber-physical systems, called “cyberizing the physical” for

specifying physical systems with computational abstractions and interfaces and “physicalizing

the cyber” for expressing abstractions and interfaces of software and network components to


61

represent physical systems’ dynamics in time [4]. Cyber-physical systems therefore are

considered as key-enabler to develop the “Internet of Things” [5][6] due to integrated complex

logics for information processing, intelligent sensors and actuators as well as their ability to

cooperate in cyber-physical systems’ networks [7].

The installation of cyber-physical systems in the production form a smart environment, called

cyber-physical production systems [8]. In this environment cyber-physical systems get

interconnected in networks using today’s ubiquitous broadband communication infrastructure.

They actively make decisions, manage processes and trigger events in order to improve for

example:

Efficient production and logistics,

Adaptive manufacturing,

Quality assurance,

Predictive maintenance,

New or enhanced business processes, and

Custom-tailored functionalities.

All activities of cyber-physical systems support the horizontal integration of participants along

the value-added chain and the vertical integration of participating stakeholders and

information technology tools.

Therefore they grant access to component-individual on line data throughout their life cycle

for example from the shop floor during production or during usage. Appropriate data storages,

infrastructures and mechanism for data access are mandatory.

3 Semantic Product Memory

The application of digital product memories to components enable the gathering of

component-individual data throughout their lifecycle [1]. Digital product memories are

miniaturized digital data storages being physically attached to concrete components. The

gathered data of a digital product memory focuses on all information which is related to the

concrete physical component and intentionally created during component’s creation and use

[2][9][10].

Digital product memories store data on the component [2]. Common technologies used for

these product memories are radio based tags like radio frequency identification transponder

(RFID) and near field communication chips (NFC) as well as two-dimensional barcodes like

the quick response code (QR code) and the data matrix [9].


62

Due to the technical limitation of these product memories, the amount of available data storage

is limited.

Data Format. An appropriate data format is needed. Digital product memories use a binary

data format [9]. This propriety data format is only readable using equivalent information

technology tools [9]. The semantic product memory, a subclass of digital product memories,

aims to provide a machine-understandable description of the digital product memory [2]. It

therefore uses semantic web technologies to describe the primitives and ontologies for the

stored data [2]. Open access to the stored data is enabled instead of restricted access as

found for the digital product memories [2].

The semantic product memory defines a modular container format. Existing digital product

memories as well as other data can be embedded [9]. The container format consist of three

sections: a header, followed by a table of contents of the block headers and multiple data

blocks [2][9].

Descriptive metadata about for example the creation date, the author or the phase of the

product lifecycle are stored in the header or the block header. For the data blocks multiple

content types are allowed like plain text, HTML, images, archives or byte streams [9]. Unique

identifier, structures and relationships are equally stored in such data blocks [10]. They are

used to create references to distributed semantic product memories or composite product

memories. Composite product memories contain semantic product memories of multiple

components. They provide an additionally infrastructure to access every single semantic

product memory [10].

Challenges. The total size of available storage has always be kept in mind while storing data

to the semantic product memory. Therefore two challenges occur: aggregation of data and

storage of life cycle information.

Distributed storage of data complicates the aggregation and update of data [10]. The semantic

product memory uses a peer-to-peer infrastructure to update and gather information [10].

Searching and updating processes therefore use requests, which are forwarded to all

neighbors [10] at a cost of time and at cost of overhead data.

As the amount of storage data is limited while storing life cycle information, semantic product

memories propose to overwrite previous data using a circular buffer [11]. Consequently

component life cycle data is only accessible for a limited time period. After this time period the

information gets lost.

4 Integrated Product Data Model

Appropriate knowledge about products and processes are crucial for successful engineering.

Due to the increasing virtualization of the product development and production planning,


63

knowledge, information and data is stored to semantic product data models. Integrated

product data models support the whole product lifecycle and enable a continuous flow of

information throughout all product life cycle phases. Such an integrated product data models

is the Standard for the Exchange of Product Model Data (STEP) as specified in the

International Organization for Standardization 10303 (ISO 10303).

Format. STEP aims to reduce heterogeneity of product model data and thus reduce the

amount of data exchange interfaces needed to integrate all information tools in product

development [12]. It therefore defines a core structure, which is specialized for every use

cases in application protocols [12].

STEP application protocols include product data such as [12][13][14]:

General management information,

Part identification,

Product structure and assemblies,

Requirements and functionality,

Geometry,

Machining form feature,

Product manufacturing information,

Product configuration,

Product status, and

Presentation.

All STEP application protocols use singular text files in the ASCII format. These text files

provide a machine- and human-readable representation of the whole integrated product data

model.

Challenges. The integrated product data model faces two relevant challenges: handling

information spread over multiple information technology tools and lifecycle interdependencies.

The integrated product data model integrates all product information. Information spread over

multiple information technology tools therefore is aggregated in one singular file. This

approach improves data access and exchange, but the actuality of high dynamic data is not

respected.

Another challenge for the integrated product data model is the representation of lifecycle

interdependencies. For every product one virtual product object is instantiated. Multiple

components physically manufactured at shop floor are then virtually represented by only one


64

singular virtual product object. All component data is aggregated within this object. A

dedicated separation between different component life cycle data and thus the source of data

is not possible even if obligatory for engineering.

Interdependencies due to specific product concepts and physical processes combination

during component’s life cycle are not explicitly recorded in the data model. Consequently the

derivation of knowledge for subsequent product development and production planning based

on these given data is restrained.

5 Integrated Component Data Model

The aim of the integrated component data model is to close the gap between product data in

the integrated product data model on the one side and component-specific data from the

semantic product data model on the other side.

The integrated component data model is an approach for the specification of the semantic

representation of the component data model. It integrates product, process, resource and

component data as well as it provides the corresponding behavior. It concerns the whole life

cycle from product idea to the component recycling or disposal.

Distinction. The integrated component data model thus extends both previously described

approaches of the semantic product memory and of the integrated product data model.

In contrast to the semantic product memory the integrated component data model is integrated

actively to all engineering processes, including the product development and the production

planning. Besides component-specific data the integrated component data model assures a

bidirectional association between component states and their behavior in the physical and

virtual world throughout the whole product and component life cycle. In contrast to the

semantic product memory this life cycle already starts before the physical creation of the

component and also includes planned life cycle events in the future of the component. To be

highlighted is the fact that the approach is not only the data model for data storage of semantic

product data, but is involved actively in the current and planned processes due to its

component-specific behavior.

Compared to the integrated product data model the integrated component data model is

focused on the component. Each component is represented by a specific instantiated object

of the integrated component data model. The identification of specific components and their

data is assured. Instead of data aggregated in a singular product data model instance, specific

data as result of concrete life cycle states and events is stored and kept in relation to the

component. Knowledge about interdependencies between product concepts and production

gets recognizable and traceable. To be more precise, a remarkable example for the difference

between the integrated component data model and the integrated product data model is the


65

storage of concrete dimensions instead of nominal dimensions with tolerance fields (see

Figure 1).

Figure 1: From product model to component model

Data Format. The approach for integrated component data model consists of core and partial

information models (see Figure 2). The core information model specifies the identification,

addressing, localization, administrative and organizational information as well as the

geometric representation [7]. Partial information models include data resulting from other

virtual and physical processes like the simulation, the manufacturing or the assembly [7]. Due

to the modular characteristics of the integrated component data model each partial model

further extends the core data model.

Figure 2: Core and partial models of the integrated component data model

Each partial information model must consequently consider formal requirements [7]:

Integrity (semantic correctness of information),

Coherence (correlation of information without transformation),

Accumulation (explicit representation of all relevant information), and

Association (derivation of implicit information) [7].


66

The seamless integration of the integrated component data model into the existing

infrastructure is currently investigated. A promising approach is the use of a federation based

on RESTles web services [15][16] and the extensible markup language (XML).

The approach for the integrated component data model consists of a variety of different core

and partial models. In the following the approach for the partial model for production planning

and its relevance in the engineering process is described.

6 Smart Production Planning

Today's production planning is driven by orders. In the smart and efficient factory of the future

components control actively or at least influence directly their life cycle phases. In the

component creation phase they control their production. They are aware of their environment

and planned life cycle states. Consequently they aim for the appropriate adoption of their

situation to reach these planned life cycle states. Thus in the factory of the future production

is no longer driven by orders, but by components.

In this context bi-directional flow of information and transparency of information is crucial for

engineering processes. Engineers of the production planning need a holistic understanding

about interdependencies in the production processes. These production processes face

component-specific production processes which are globally planned for product series, but

individualized upon production. Each component is individually produced depending on

environmental influences like customization, production bottlenecks or machine breakdown.

Consequently production planning is specific for each component. Interdependencies are

manifold, but recorded to component-specific life cycle histories in the integrated component

data model.

In this paper an approach, called smart production planning, is presented. Smart production

planning uses the integrated component data model to increase the efficiency in the resource-

efficient factory of the future by the front-loading of detected interdependencies between

production planning and shop-floor states. Thus it creates a data base for deriving the required

holistic understanding about the production processes.

Example. The integrated component data model specifies partial information models for the

wear of tools in milling machines at shop floor and resulting producible tolerances on the one

hand. On the other hand it equally specifies numerical control (NC) code as result of the

computer aided manufacturing (CAM) as well as the corresponding tolerance fields chosen

during the product development.

Creating relationships between this information and its use in the smart production planning

allows the improvement of the tool wear exploitation. The smart production planning therefore

enables production planner to schedule milling operations with the respect to the planned

tolerance fields. Operations with respect to narrow tolerance fields are scheduled first,


67

operations with wider tolerance fields are scheduled later. The milling tool is exploited at the

maximum of its manufacturing capacity. The efficiency of the production process is increased.

7 Conclusion

The integrated component data model is an approach for the specification of the semantic

representation of component data. It specifies core and partial information models to integrate

product, process, resource and component data. The integrated component data model

further extends the semantic product memory in terms of life cycle support, represented data

and behavior. It equally extends the integrated product data model in terms of component-

specific data storage and retrieval.

The use of the integrated component data model within the production planning makes the

process smart. Integrated data enables production planner to derive interdependencies

between product concepts and production processes. In this paper an example for the

improvement of the resource-exploitation of milling tools is given. There the integrated

component data model is used to reschedule milling operations in respect to concrete tool

wear at the shop floor in relation with producible tolerance fields and approved tolerance fields

of the product development.

Upcoming work in the research project is the further specification of the integrated component

data model and its integration to existing product development and production planning tools.

A promising approach therefor is the use of a federative architecture as proposed in the

federative factory data management [16]. Additionally research projects analyze the use of

the integrated component data model in other engineering processes and its impact on the

engineering processes [17].

8 References

[1] Broy, M.; Kagermann, H.; Achatz, R. (eds.): Agenda Cyber Physical Systems – Outlines

of a new research domain. acatech – National Academy of Science and

Engineering, München, 2010.

[2] Wahlster, Wolfgang: The Semantic Product Memory: An Interactice Black Box for Smart

Objects. In: SemProm – Foundations of Semantic Product Memories for the

Internet of Things, Wolfgang Wahlster (eds.). Springer-Verlag, Berlin, Heidelberg,

2013, pp. 127-148.

[3] Eigner, Martin; Anderl, Reiner; Stark, Rainer: Interdisziplinäre Produktentstehung. In:

Smart Engineering – Interdisziplinäre Produktentstehung, Reiner Anderl, Martin

Eigner, Ulrich Sendler, Reiner Stark (eds.). acatech Diskussion, München, 2012,

pp. 7-16.


68

[4] Lee, Edward A.: CPS Foundations. In: Design Automation Conference ’10, Anaheim,

California, 2010, pp. 737-742.

[5] Broy, Manfred (eds.): Cyber-Physical Systems. Innovation durch Software-intensive

eingebettete Systeme. acatech DISKUTIERT, Springer, Heidelberg 2010.

[6] Acatech (Hrsg.): Cyber-Physical Systems. Innovationsmotor für Mobilität, Gesundheit,

Energie und Produktion. acatech POSITION. Springer, Heidelberg, 2011.

[7] Anderl, Reiner; Strang, Daniel; Picard, André; Christ, Alexander: Integrated Component

Data Model for Industrie 4.0 – Information Carrier for Cyber-physical Production

Systems. In: ZWF – Zeitschrift für wirtschaftlichen Fabrikbetrieb (1-2/2014), 2014,

pp. 64-69.

[8] Kagermann, Henning; Wahlster, Wolfang; Helbig, Johannes (eds.): Recommendations for

implementing the strategic initiative INDUSTRIE 4.0.acatech – National Academy

of Science and Engineering, Frankfurt/Main, 2013.

[9] Horn, Sven; Claus, Alexander; Neidig, Jörg; Kiesel, Bruno; Hansen, Thorbjørn; Haupert,

Jens: The SEMPROM Data Format. In: SemProm – Foundations of Semantic

Product Memories for the Internet of Things, Wolfgang Wahlster (eds.). Springer-

Verlag, Berlin, Heidelberg, 2013, pp. 127-148.

[10] Horn, Sven; Schennerlein, Barbara; Pförtner, Anne; Hansen, Thorbjørn: Distributed

Digital Memories. In: SemProm – Foundations of Semantic Product Memories for

the Internet of Things, Wolfgang Wahlster (eds.). Springer-Verlag, Berlin,

Heidelberg, 2013, pp. 127-148.

[11] Neidig, Jörg; Preißinger, Jörg: A SEMPROM Use Case: Maintenance of Factory and

Automotive Components. In: SemProm – Foundations of Semantic Product

Memories for the Internet of Things, Wolfgang Wahlster (eds.). Springer-Verlag,

Berlin, Heidelberg, 2013, pp. 363-380.

[12] Anderl, Reiner; Trippner, Dietmar (eds.): STEP – Standard for the Exchange of Product

Model Data. B.G. Teubner, Stuttgart, Leipzig, 2000.

[13] http://www.ap242.org – last visited on 2014-07-18

[14] International Organization for Standardization: Application protocol: Product life cycle

support. In: International Standard ISO 10303-239:205 - Industrial automation

systems and integration — Product data representation and exchange, Geneva,

2005.

[15] Picard, André; Anderl, Reiner; Schützer, Klaus: Controlling Smart Production Processes

Using RESTful Web Services and Federative Factory Data Management. In: 14th


69

Asia Pacific Industrial Engineering and Management System, 2013, Cebu,

Philippinen.

[16] Picard, André; Anderl, Reiner: Smart Production Planning for Sustainable Production

based on Federative Factory Data Management. In: Proceedings of TMCE 2014,

Imre Horvath, Zoltan Rusak (eds.). Budapest, 2014, pp. 1147-1156.

[17] Strang, Daniel; Galaske, Nadia; Anderl, Reiner: Beschreibungsmethode für die

Repräsentation cyber-physischer Produktionssysteme. In: Entwerfen, entwickeln,

erleben, Ralph Stelzer (eds.). TUDpress Verlag der Wissenschaften GmbH,

Dresden, 2014, pp. 13-26.


70


71


72


73


74


75


76


77


78


79

80

Marcio Weichert

Marcio Weichert é, desde maio de 2012, coordenador do Centro

Alemão de Ciência e Inovação São Paulo (DWIH-SP) e coordenador

do programa acadêmico e científico do Ano Alemanha + Brasil 2013-

2014. Bacharel em Comunicação Social pela Universidade Federal

Fluminense (UFF), em Niterói-RJ, fez carreira como jornalista no Rio.

Atuou nos jornais O Dia e O Globo, na Bloch Editores, bem como em

assessorias de comunicação e imprensa e como editor independente.

Na Alemanha, trabalhou na Deutsche Welle, a emissora internacional

da Alemanha. De volta ao Brasil, fez pós-graduação em Gestão

Estratégica na Universidade Cândido Mendes (UCAM) e tornou-se

assessor de marketing e comunicação do Serviço Alemão de

Intercâmbio Acadêmico (DAAD), função que exerceu de fevereiro de

2006 a maio de 2012.

[email protected]

Centro Alemão de Ciência e Inovação São Paulo

Projeto de iniciativa do Ministério das

Relações Externas da Alemanha, o Centro

Alemão de Ciência e Inovação São Paulo

(DWIH-SP) nasceu em 2009, tendo

funcionado provisoriamente na Câmara de

Comércio Brasil-Alemanha até 2011. Em

fevereiro de 2012, foram inauguradas suas

atuais instalações, com oito escritórios e uma

sala de reuniões. Sua principal missão é

servir de ponto de encontro e ponte para

instituições e profissionais dos meios

acadêmico, científico, de inovação e de

fomento da Alemanha e do Brasil. Três

agências de fomento, cinco representações

de instituições de ensino superior alemãs e a

sociedade de pesquisa aplicada Fraunhofer

estão reunidas no DWIH-SP. As instituições

de ensino e pesquisa alemãs buscam

cooperar com a indústria no Brasil, sejam em

projetos de pesquisa e inovação, sejam na

capacitação profissional.

81

Cooperação com Universidades Alemãs:

Oportunidades para a Indústria no Brasil

Resumo

O Centro Alemão de Ciência e Inovação São Paulo (DWIH-SP) busca, entre outros

objetivos, aproximar instituições de pesquisa alemãs e empresas no Brasil para

cooperações de vários tipos. A palestra informará as opções. Como coordenador da

programação científica da Temporada Alemanha+Brasil 2013-2014, o DWIH-SP

estimulou eventos com este objetivo. A palestra apresentará um resumo das iniciativas

em andamento.

Palavras chave

Ciência; inovação; pesquisa; intercâmbio; cooperação; fomento; universidade-empresa.


82


83


84


85


86


87


88


89


90


91


92


93

94

Dr. Andreas Romberg

Dr. Romberg was born in 1962 and studied mechanical engineering

at the University of Kaiserslautern, Germany, where he received his

Diploma in 1988. In 1992 he earned his Dr.-Ing. degree in production

engineering and business organization also at the University of

Kaiserslautern. Dr. Romberg gained profound lean and managerial &

leadership knowledge during his industrial career working for several

automotive suppliers within production and as a plant manager. He

also implemented different lean systems with focus on production and

supply chain within this period. Since 2003 he has been working as a

business consultant and management coach for Staufen AG. He is

Head of Business Unit „Innovation and Product Development“. His

core competencies are: Lean Management in Innovation and Product

Development; Project Management in multi-project environments;

Shopfloor Management; Implementation of holistic Value

Management Systems; Ramp-up Management. Dr. Romberg has

authored or co-authored 2 books and about 20 papers.

[email protected]

Staufen AG

Staufen AG is one of the leading Lean

Management consultancies in Germany. As

"Partner on the way to Top-Performance", the

internationally operating consulting house

supports companies in optimising their value

creation and management processes as well

as increasing the efficiency of their innovation

and product development processes.

Furthermore, the consultants develop

concepts to manage crisis situations as

turnaround or interim managers. With the

Staufen Academy, the consultancy also

offers certified practice-oriented training.

More than 200 employees look after

customers on site from offices in Germany,

Switzerland, Italy, Poland, Czech Republic,

Hungary, Slovakia, China and Brazil. The

recent Lünendonk® 2013 trend study on

"Performance - growth strength of

management and IT consultancy companies"

ranks Staufen among the consulting houses

with the strongest growth over the last five

years.

95

Frontloading is a Key Success Factor and a

Basis for Efficiency and Effectiveness of NPI

Projects

Abstract

There is a saying in Germany: “Tell me how you are starting a project and I will tell you

how it will end up!” Nothing has been more confirmed in our 10 years of project experience

within Lean Development than this saying. Most of the projects we saw life in the multi

project environments of our customers have been started without taking advantage of

better input up front the project.

Frontloading anticipates problems in a project before they even can occur. That kind of

early risk assessment and advanced problem solving helps to reduce risks. Good

Frontloading starts with the input of a sufficient customer/ market specification data to

make sure that especially engineering has understood the content as well as clarified the

priority of conflicting features and functions. Frontloading, from the organizational point of

view, is a very cross-functional approach to collect all the ideas, problems and other

relevant issues of all NPI (New Product Introduction) process related functional

departments. Even the inclusion of key suppliers is helpful. The outcome is a whole set of

solutions partly based on already existing solutions which helps to reduce effort, risks/

quality issues and costs. Frontloading also enhances the team-building process. The

common assessment of specifications, anticipation of possible problems and risks in early

stages of the project creates a common awareness of the key success factors in the

project. This contributes to a higher commitment of team members to the project.

The lead-time of a well-moderated frontloading process from our experience lasts about a

third of the whole lead-time of a product development process. Our recommendation: Have

the courage to spend this time because it will help you to save more time during following

realization and validation phase.

Keywords

Set based concurrent engineering; point based engineering; cross-functional approach;

maturity model for predevelopment; early risk assessment; efficiency; effectiveness; lean

development system.


96

1 Introduction

As the Chaos Institute considered in its Chaos Report 1994 [1], 82% of all projects do not

achieve the project targets and/ or customer requirements.

Figure1: Achievements of objectives of projects [1]

That means projects do not complain with planned deadlines (performance delivery) or they

exceed the planned budget and product costs. Sometimes projects compromise in terms of

quality by missing certain functions and features. This has multiple root causes. One of the

major root cause is missing cross-functional frontloading. That means we miss the chance for

projects to:

Verify the completeness of all necessary and sufficient input data about specification and

frozen objectives especially of the customers’ side.

Use common and cross-functional knowledge about the project tasks including the

knowledge of key suppliers, which supplement the projects with their core competencies.

Re-use standards for parts, modules and technologies; avoid re-inventing the wheel.

Verify the maturity of technologies coming out of pre-development processes. Lack of

maturity of technologies often leads to unnecessary and costly loops within product

development projects.

As shown in Fig. 2 the above-mentioned circumstances of missing or bad frontloading first of

all normally cause high effort in late project phases. And the second aspect of this additional

effort is that this is normally a not planned effort.


97

Figure 2: Missing or bad frontloading leads to high effort in second half of projects [2]

An additional amplifying effect occurs when imbalanced project pipelines meet missing or bad

frontloading. We more frequently recognize that the project environments of companies have

most of their projects in late phases due to the fact that engineering has to support production

facilities ramping up the new products longer times after SOP. Compared to that there are

only a small number of projects in early phases. Combined with high-unplanned project effort

within the late phases companies are often firefighting and have no time/capacity to frontload

new projects. This very early starts to become a vicious circle.

Figure 3: Imbalanced project pipeline meets projects with high effort in late phases

To get out of that situation companies must start with additional effort to get frontloaded

projects. Step by step the red curve in Fig. 4 will move to the direction of the blue curve.

Frontloaded projects have the most of their effort in the first half of the NPI process and this

effort is planned.


98

Figure 4: Ideal state: Frontloaded projects have most effort in the first half of the project

2 About Frontloading

It is common knowledge that (product-) costs are determined by Engineering (Fig.5) in early

phases of NPI processes [3] hence it is strictly recommended to take care about well-done

frontloading.

Figure 5: The determination of costs takes place in the first phases of NPI process [3]

Frontloading supports effectivity and efficiency within NPI process; typical characteristics of

good frontloading are shown in Fig. 6:


99

Concept phase/ frontloading lasts round about 1/3 of a whole NPI project.

Clarifies all necessary and sufficient information about project tasks and objectives, such

as market and product specification.

Frontloading is a cross functional discipline; all necessary functional departments are

represented within the frontloading team to support the process with their knowledge.

The frontloading team is working highly integrative within cross-functional workshops

moderated by the chief engineer (= project lead).

The cross functional workshops ideally are taking place in a big project meeting room

(= OBEYA) with visual management of all relevant information.

After clarification of the product specification, the team is working in sub phases on

functional models and product architecture with a set based approach.

The product concept will be narrowed by the following sub phase “concept aggregation

& evaluation”.

Frontloading ends up with a clear written system specification, which contents all

information how the new product will be realized. This specification ideally is frozen after

that milestone (M3).

Figure 6: Characteristics of frontloading [4]


100

Comparing traditional NPI processes/ projects with a Lean NPI process using frontloading it

is possible to reduce lead-time > 25 … 30%. Even though spending more time as usual in the

first phase due to a longer frontloading, the lead-time of the whole project will be shortened

by the reduction of realization and validation phase (see Fig. 7). That is the result of permanent

reconciliation within the cross-functional workshops discussing and solving future problems

before they occur and avoiding risks normally followed by time and budget consuming process

loops. Doing the things right – right from the start.

Figure 7: Frontloading: The effect of effective collaboration [4]

3 Set Based Concurrent Engineering

Set based concurrent engineering is a process approach, which in contrast to the traditional

point based product development requires multiple design solutions, outlining how Japanese

companies gain advantage in delaying design decisions by relying on sufficient knowledge

[5].

As illustrated in Fig. 8, the point based approach limits the design space upfront, hence

providing less flexibility to adjust design solutions among the different product development

functions. Set based approach on the other hand enables product development functions to

explore design space and converge to an optimum design solution during the set narrowing

phase [6]. Set based concurrent engineering is a knowledge intensive process that comprises

the communication, trade-off and narrowing down a set of potential design solutions whilst

proceeding in product development [7].

Once a reasonable solution has been examined (end of narrowing phase) the cross-functional

fine-tuning and adjustment starts within a kind of problem solving or problem correction phase.

In case of an unresolvable problem the team can with a small step back choose an alternative

solution out of the former solution set.


101

Figure 8: Point based vs. set based engineering [6]

In contrast to that within the point based approach every pretended step further is very often

also combined with bigger changes and adjustments which are time consuming and often very

costly.

Figure 9: Effective among others means “using existing knowledge and experiences” [8]

Providing the right knowledge to the right time is the key issue within Lean Development

processes. A three dimensional model of knowledge management has been emerged to be

vital for effective product development [8].

Horizontal dimension (X) symbolizes that knowledge is required to sequentially proceed in the

product development process. For example, in the design (D) function the knowledge acquired

in the concept (D1) phase, such as new customer requirements, is provided to the detail


102

design team in order for it to be realized in the new design. Vertical dimension (Y) exemplifies

that knowledge needs to be obtained or shared within other functions in the product

development process. This can include sharing knowledge concurrently between the

manufacturing (M) and design (D) function during the concept phase in order to assure that

manufacturing feasibility is considered at an early stage of the product design process.

Previous projects dimension (Z) embodies knowledge a company has acquired in the past.

For instance, during validation (D3) in the design (D) function, the product development

engineer retrieves proven test configurations from previous projects in order to initiate the

validation process [8].

Representing, sharing and shifting the knowledge in a way that it can be easily adapted is the

challenge. Because a big part of the knowledge is tacit knowledge driven by lots and lots of

experience, which is actually difficult to represent.

Experienced driven knowledge also means that companies have to handle lots and lots of

data. There are two interesting methods to represent such kind of data.

One shown in Fig. 10 are “trade off curves” which is already used for a long time by TOYOTA

[9]. Trade off curves are showing correlations of different impacts on certain product design

issues. The example in Fig. 10 shows some correlations for the design of a car seat structure

[8] such as product unit cost vs. production volume or material cost vs. weight and so on. Such

knowledge representation combine “big data” volumes on some easy to read, handle and

maintaining graphs.

Figure 10: Examples for trade of curves [8]


103

The second (Fig. 11) shows how to capture and provide knowledge coming from A3 problem

solving processes.

Figure 11: A3 problem solving template completed by 2 steps to capture and provide knowledge [8]

Fig. 12 shows an example of an A3 problem-solving template completely filled. The knowledge

is captured within decisive design rules or design recommendations to be applicate for similar

future designs in a definite process phase.

This knowledge could easily be provided within a database.

There are many other partly well-known methods to increase effectivity in product

development processes such as: QFD, value analysis, value engineering, design for

manufacturing and assembly, value management for complexity etc. These methods are also

applicable within frontloading – the earlier the better.

Efficiency means “do the things right”. What causes the effect of efficiency within frontloading?

As shown in Fig. 13 the effect of being efficient can be recognized by a tremendous lead-time

reduction. Good frontloading within product development enables us to switch from a

sequenced NPI process to a highly parallelized NPI process.

Throughout an intensive cross-functional discussion on the best product design we can find a

very good reconciliation between the different relevant functions. Everyone has a clear picture

what to do and can work in parallel with all the other functions.


104

Figure 12: Example for a completed A3 problem-solving template [8]

Figure 13: Efficiency effect caused by frontloading

In case of any problem, open issue or question the team together will work on a cross

functional solution supported by OBEYA and visual management as the major tools.

So while Engineering & Development is working on the best product design the Process

engineering is working on the ideal processes. Even the market development & introduction

can be work out by Product Management in parallel.


105

4 Summary

Frontloading as a part or a principle of Lean Product Development is a process approach

narrowing a final product design out of a set of feasible solutions (design space) which have

been worked out within a cross functional team supported by the application of some methods,

tools and techniques. By using an OBEYA, project-/ process related communication within the

team is parallel and almost free of interfaces. Visual management supports the product

development process and team discussion with all necessary and sufficient information such

as calculation data, BOMs, sample parts, competitor information and parts etc.

The application of frontloading happens in very early stages of a product development process

long before the release of major investment, tooling issues and so on. Frontloading from the

beginning also focusses strongly on external and internal customer satisfaction (“doing the

right things” effectivity) as well as short lead times (“doing the things right” efficiency).

Resources are also able to invest their capacity because they are not planned out by 100%

but only 75 … 80% of their theoretical available capacity. So they are able to put in their

knowledge.

This from our experience helps to reduce lead-time tremendously by avoiding a lot of waste

within traditional approaches and processes. Some of our customers reduced lead-time up to

60% and they improved their project throughput up to 40% by the same amount of resources!

Normally due to the fact that lead-time reductions also cause a higher project throughput the

innovation rate also increases from the lower one-digit ranges up to 25 … 30%. Innovation

rate thus describes the part of the annual turnover with products younger than the half of their

typical life cycle.

5 References

[1] Chaos Report (1994). The Standish Group

[2] Romberg, A. (2010). “Schlank entwickeln, schnell am Markt - Wettbewerbsvorteile durch

Lean Development“. Log-x-Verlag.

[3] Lindemann, Mörtl (2008): „Kostenmanagement in der Produktentwicklung“ (= “Cost

management in product development“) Chap.1, Part 1

[4] Romberg et al (2013): “Lean Development Trainer”; Training material Staufen AG.

[5] Ward, A. C. (2007), Lean product and process development, Lean Enterprises Inst. Inc.

[6] Ward, A., Liker, J. K., Cristiano, J. J. and Sobek, D. K. (1995), "The second Toyota

paradox: How delaying decisions can make better cars faster", Sloan management

review, vol. 36, pp. 43-43


106

[7] Sobek, D.K.; Ward, A. C.; Liker, J. K.(1999). “Toyota’s Principles of Set-Based Concurrent

Engineering”. MIT Sloan Management Review.

[8] Maksim Maksimovic (2013), "Lean knowledge life cycle framework to support lean

product development”, PhD Thesis, Cranfield University, England.

[9] Liker, J. K.; Morgan, J. M. (2006). “Toyota Product Development System”. Taylor &

Francis Inc.


107


108


109


110


111


112


113


114


115


116


117

118

Dr.-Ing. Peter Binde

Dr. Binde was born in 1970 and studied mechanical engineering at the

Universität Darmstadt, Germany, where he received his diploma in

1997. Towards the end of his studies, he began working as a

freelancer consultant in CAD/CAE. He received the Dr.-Ing. degree in

Mechanical Engineering at the Universität Darmstadt in 2004. In 2005

he founded Dr. Binde Ingenieure, Design & Engineering GmbH, a

company that specializes in CAE consulting all around the Siemens

NX System. The company grew steadily and now employs engineers

who are all experts for product simulation. The fields of expertise are

Structural Mechanics, Rigid Body Mechanics, Fluid Mechanics,

Thermodynamics, Electrodynamics and the integration of all this in

PLM. 2009 he established a code for electromagnetic analysis

integrated in the NX System. For this he performed government-

funded research projects together with the Universität Darmstadt and

University of Liège, Belgium. From 2008 to 2012, he taught Numerical

Structural Analysis and Multi-Body Dynamics at the RheinMain

University of Applied Sciences. Since 2011 he leads the German CAE

Special Interest Group in the PLM Connection, a user group which

meets twice a year on the topic of simulation.

[email protected]

Dr. Binde Ingenieure, Design & Engineering GmbH

119

Recent Approaches of CAD/CAE Product

Development. Tools, Innovations,

Collaborative Engineering

Abstract

In this paper, the latest approaches in the field of CAD-CAE product development are

presented, as they are applied in industry. Innovations in the software tools are shown

such as they arise from multi-physical simulation technologies. The implementation of

these processes in collaborative engineering is discussed, such as the relation between

OEM and supplier.

Keywords

CAD; CAE; Simulation; Product development; Collaborative engineering.


120

1 Introduction

In former times integrated CAD analysis tools were limited to basic applications only. These

were mainly linear Finite Element Analysis (FEA, FEM) types for strength analysis and Multi

Body Dynamics (MBD) analysis for kinematics.

The focus of these tools was the design- and not the analysis-engineer. Goal was easy setup

of analysis models and fast responses to enable designers for A-B comparisons and

decisions. Designer’s analysis tasks are typically characterized by less abstraction

techniques, so his 3D CAD models should not be modified or idealized very much for analysis.

Consequently meshing for FEA was performed by simple tetrahedral elements which can well

be automatically created on solid geometry. These element types have been designed in such

a way that for the basic discipline of linear statics even poor quality element-shapes already

led to relatively accurate results. Therefore adaptive meshing strategies and high-grade

polynomials were developed in FEA shape functions. Still a problem for designers is correct

validation of strength results, so decisions about strength are usually not done by them rather

by analyst- or measurement engineers.

Analysts in former times never used CAD integrated simulation tools, because of their usual

need for more abstraction and complexity in geometry and physics. Missing functionalities

relate in particular to possibilities for non-linear simulation (contact, material and geometry),

advanced material laws, advanced meshing techniques with shells, beams and hexahedral

elements and the coupling of different simulation methods. Additionally in many cases there

exist self-made software codes for individual problems which must be coupled or integrated

to FEA or MBD systems to be efficient.

One more limitation in former systems was the focus on mechanical engineering only. Other

engineering disciplines like electrical-engineering, control-engineering and system simulation

were not supported.

The situation in today’s large CAD software vendors is characterized by the fact that they more

and more must satisfy needs coming from big OEM customers such as automotive and aircraft

industries. This leads to large software vendors take over smaller companies that have

specialized technologies and try to include these technologies into their major system.

Competition around these specialized technologies is running and the question how to

integrate all these technologies efficiently into the main system becomes crucial.

That’s why more and more specialized high-end CAE software becomes integrated into major

CAD systems. Analysis engineers now more and more do find their special functionalities that

were missing in the past. This is one focus in this article.


121

Figure 1: Disciplines integrated in CAD/CAE systems

Elementary disciplines that are integrated in most major CAD/CAE systems today are

Strength Analysis computing displacements, stresses and strains through FEM,

Dynamic Responses of Structures for free or forced vibration effects by use of FEM,

Multibody Dynamics taking into account rigid body motion (MBD) to compute

displacement, velocity, acceleration, joint-force and

Thermal Analysis for thermal conduction, convection or radiation by FEM.

The following two disciplines are more advanced and currently at the beginning of becoming

popular in major systems:

Computational Fluid Dynamics (CFD) for pressure, velocity, turbulence through use of

Finite-Volume-Method (FVM) and

Electromagnetic Analysis (EM) for forces, eddy-currents and field-strengths computed

by FEM.

Stand-alone tools of course are there and are powerful in their respective fields. But the

challenge today for OEMs is the utilization of integrated CAD/CAE software that does allow

for adaption to needs. Stand-alone tools must fit in this structure. So interfaces play an

important rule.

Backbone for all of this is Product Data Management (PDM), what in area of CAE turns out

as Simulation Data Management (SDM).

2 Multiphysics Solutions

While in the past solutions for those elementary disciplines mentioned above have been in

focus today and in future there are more and more solutions for coupled problems needed.

Following we describe some main types of those solution types, how they can be performed

and in which industrial applications they are used.


122

2.1 Thermal / Structural one Way

One way thermal structural coupled analysis does first compute for temperature fields and

then apply those temperature fields as loads to structural models. This is needed in all fields

of strength analysis cases where thermal expansions plays a role. An example are motor

housings.

Since mapping of temperature data from a thermal mesh to a structural mesh through

interpolation methods is not very complicated this analysis type is not difficult to perform. It

can be done in most integrated CAD/CAE systems today.

2.2 Thermal / Structural two Ways

A two way coupled thermal structural analysis is much more sophisticated. Temperature loads

lead to structural deformation similar to the simple one way coupling described above. But the

two way approach takes into account that deformed models may lead to different thermal

conditions. This particularly appears if there exist contacts that – if closed - transfer thermal

fluxes. Example applications for this are screwed container seals in nuclear plants.

This type of analysis is much harder to perform since iterations must be carried out and some

convergence criteria must be controlled. Also meshes are needed that are good for the

thermal as well as for the structural part. Because of those many iterations result mappings

between different meshes should be avoided.

In todays integrated CAD/CAE systems this type of analysis is beginning to take place.

2.3 Thermal / Fluid two Ways

The combined analysis of thermal and fluid is carried out separately in rigid body regions and

in fluid regions. At all interfaces there must be solved for heat transfer conditions. Example

applications are cooling of electronic systems.

The analysis in rigid bodies is of simple thermal type. In fluid regions Navier-Stokes-Equations

are solved by CFD. Since CFD is already an iterative solution that takes temperatures into

account an application of rigid body temperatures is not very much more complicated.

Several systems today do have availability for this analysis type. But still this is not common

for most of them.

2.4 Fluid / Structural one Way

Forces and pressures arising from fluid lead to deformations. These effects are taken into

account by first analyzing for flow, capturing forces on walls from pressure results and


123

transferring them to following structural analysis. Again mapping between different meshes

must be carried out. Applications are stationary aircraft wing investigations.

Only some of the integrated CAD/CAE systems allow this analysis type.

2.5 Fluid / Structure Interaction (FSI)

The case FSI is fluid structural coupling in two ways. Fluid forces lead to deformations and

those deformations lead to different fluid conditions. Applications are aircraft wings and turbine

blades braking due to FSI oscillations.

FSI type analysis is currently not well established in the major integrated CAD/CAE systems.

2.6 Electromagnetic / Structural one Way

Electromagnetic forces, for example Lorentz-forces, are computed in the EM solver and

transferred to structural models to be solved for deformation, stress and strength. Applications

are minimum air-gap studies in electric motors. This analysis type and all following regarding

to EM, are possible in few systems only.

2.7 Electromagnetic / Thermal one Way

Losses that result from electromagnetic eddy-currents and hysteresis effects, are computed

in EM solvers and then used as thermal loads in following temperature studies. Application is

electric motor thermal analysis. Few systems only allow this type of analysis.

2.8 Electromagnetic / Thermal two Ways

Again losses are computed by EM and used to find temperature fields in the second step. But

now those temperatures lead to different material-properties and back influence the EM result.

Particularly the electric conductivity in electro-sheets of motors varies heavily with

temperature. So common applications are electric motors again. Again, only few systems

allow this type of analysis.

2.9 Control-System / Dynamic Response

For example in machine tools, vibration behavior is improved with controller circuits. For

simulating these effects either control system models must be integrated into FEA or FE

models must be integrated into control system models. Few systems only allow this type of

analysis.


124

3 Solver Languages

All major FEA solvers today provide thousands of commands that allow analysis of very

special problems. In common cases only small percentages of all commands are used in the

analyst’s daily work. User interfaces allow easy finding necessary commands for the daily

work.

One recent approach addresses general ways how to implement new solver technologies in

CAD/CAE systems. The need for this comes from the fact that more and more special

technologies must be implemented in large CAD/CAE systems. This method utilizes so called

solver languages and a neutral language for solvers. Currently this is developed for FEA

solvers, but in future it may be available for other types like MBD too. Key idea is that all input

data for any FEA solver can be classified by the following set of objects:

Element Quality Checks: Special quality checks for the considered solver.

Solution Class: Description of all solutions that characterize the solver, for instance

Thermal or Structural or Electromagnetic.

Sections: One-dimensional elements may need various sections.

Elements: The various element types a solver can handle.

Physical Property Tables: All physical properties like material data.

Modeling Objects: Additional data blocks.

LBCs: Loads, boundary conditions, constraints and related data.

Solution: Detailed description of the physical solutions that the solver can perform.

By use of these solver languages companies can implement their own solvers with specialized

capabilities into the commonly used CAD/CAE system. This increases collaboration

effectiveness between different analysis groups in companies.

The software vendor has another advantage: He can easily and fast implement new solver

technologies into the CAD/CAE system.

4 Master Model Approach

While simple analysis types – linear FEA and MBD Kinematics - became integrated into CAD

systems some basic techniques were developed that allowed associativity between CAD and

CAE data. These techniques resulted to be very successfully and today are still exploited by

CAD/CAE systems. Of course the master model approach is used there basically. This means

that all data for downstream processes is held in separate files being connected to the CAD


125

master geometry. If the CAD geometry changes everything can update. Also this gives the

basic possibility for engineers working concurrently on a digital product.

5 Designer / Analyst Collaboration

Next basic approach is the idea that CAE objects are linked to CAD objects. CAE Objects

describe the CAE problem, such are forces, meshes, boundary conditions, material properties

and others for FEA and similar links, joints, drivers, sensors for MBD. CAD objects are faces,

edges, bodies and vertexes. While these CAE objects are connected to CAD objects fully

automatic updates after geometry changes are possible. As an example you can imagine a

CAE force object. The force object stores its magnitude and direction. The direction may be

set to perpendicular to a CAD face. So if the CAD face changes because of designers work

this force will update to a new direction. The following figure shows an example of MBD objects

referencing CAD objects.

Figure 2: Associativity CAD to CAE

One newer approach supports concurrent engineering between CAD designers and analysts.

In general also analysis engineers need to modify CAD models. Commonly there must be

done simplifications, defeaturings, midsurfaces and other modifications. Therefore they need

to modify CAD geometry but it is not allowed to modify the CAD master. So this approach

offers an additional CAD model, we call it idealized CAD model, which is placed between the

origin CAD master and downstream CAE files. Associativity between the origin CAD master

and this idealized CAD model must be given by using associative geometry links. By this way

changes in the CAD master lead to updates of CAE data. Nevertheless CAE engineers do

have possibilities to modify geometry.


126

Figure 3: Idealized file and its integration in the CAD, CAE file structure

6 Design-Embedded Analysis / CAE-Experts Collaboration

A further approach for the support of concurrent engineering between design- and analysis

engineers is given by the following. There are CAE software tools used by design-engineers

and there are other tools used by analysis engineers. Both must have access to the CAD

master so in the first step they simply can be placed parallel in the master model approach.

But there are one or two additional data exchanges necessary because design-engineers may

want to transfer their CAE models to analysis engineers to make it possible to see and check

what is done there and maybe to use parts of it in more sophisticated simulations. This is the

first additional data exchange and the second one, which is optionally only, allows improved

models from analysis engineers to be back transferred to design engineers. These exchanges

between design- and analysis engineers can be realized in various ways. An optimal solution

would be if they both worked with exactly the same software, so all files could be shared. A

good compromise is to use the same solver, so data exchange can be performed via solver

input files. In many practical cases today there is still no digital data exchange between the

two groups, models are set up multiple times and communication works via telephone or

meeting-sessions.


127

Figure 4: Design embedded Analysis / Analysis Experts Collaboration

7 Data Management

The management of analysis and simulation data aims to integrate analysis and simulation

results in the workflow of virtual product development. For this purpose, this information is

embedded in a PDM environment. Background information can be found in the

recommendation [SimPDM]. An overview of the results of the SimPDM project group provides

[Anderl3].

Most manufacturing companies today face the challenge of having to develop faster and more

complex products. Design and simulation play a key role for the evaluation of product

development results. What is new for many engineers is that simulation is gaining an

increasing importance and for a higher development efficiency its integration with 3D product

modeling is a critical success factor.

This linkage problem is characterized by the following properties [AnderlBinde1]:

Personnel separation of modeling from the analysis,

Many different CAE software systems,

Many analysis variants,

Lack of relationship of CAD to CAE models,

Lack of process orientation,

Inadequate data protection,

Insufficient supplier integration.


128

In our following discussion, we restrict ourselves to the solutions of the PLM system

Teamcenter from Siemens Industry Software GmbH and its CAE expansion modules, which

are known under the name of Teamcenter for Simulation. These solutions now offer solutions

for some of the above-mentioned problems or at least approaches to overcome them.

If in Teamcenter a standard CAE analysis is performed, the native files (Figure 3: Simulation,

FEM and Idealized) are assigned to the corresponding Item Revisions (CAEAnalysis,

CAEModel and CAEGeometry) as references (see next figure). These CAE Item revisions can

be revised independently. In addition, data is automatically provided with relationships. This

data model and the relations - somewhat simplified - are shown in the following figure [TCSim].

Figure 5: CAE Data Model used in Teamcenter

The relations have the following meaning:

TC_CAE_Defining: Therefore relationship can be traced, which CAEModelRev is used

by the CAEAnalysisRev, so which meshing is computed with a SIM file. For a

CAEAnalysisRev there can be only one CAEModelRev, because there can be only one

mesh for the computation.

TC_CAE_Source: This relationship indicates from which item revision a model has been

created, so what item revision was the source. It can be defined between CAEModelRev

and CAEGeometryRev or between CAEGeometryRev and CAD Master revisions. In

case that no idealized file is used, this relationship may also exist between

CAEModelRev and CAD master. There can be only one source at a time.

TC_CAE_Target: This relationship documents for which CAD-Master-Revision each

CAE-Item-Revision applies (from CAEModelRev to CAD -Master-Revision, from

CAEGeometryRev to CAD-Master-Revision). There may exist several TC_CAE_Target


129

relationships in parallel. An example of multiple parallel TC_CAE_Target relations is the

following: The simulation for a green part shall have validity for the identical blue, yellow

and red parts.

Using this data model, the desired relationships between CAD and CAE are now available.

Thus it can answer the questions: "Which CAD model belongs to which FEM model?" and

“Are there new CAD revisions for my FEM model?”. Also the problem of a high variety of

analysis variants is addressed. Additionally, these relations allow to automate approval

processes between design and analysis.

8 Literature

[SimPDM] ProStep iViP Recommendation “Integration of Simulation and Computation in a

PDM Environment (SimPDM)”. PSI 4, Version 2.0 2008

[Anderl3] Anderl R./Grau M./Malzacher J.: SIMPDM – a harmonized approach for the

strategic implementation of simulation data management. NAFEMS World

Congress 2009

[Anderl4] Anderl R./Malzacher J.: SimPDM – Simulationsdatenmanagement-Standard nach

Maß. In: CAD CAM, Nr.1-2, 2009, Pg. 38-41

[AnderlBinde1]

[TCSim] Training documentation for Teamcenter for Simulation. Siemens PLM Software.

130

Dr. Jorge Vicente Lopes da Silva

É doutor em Engenharia Química, mestre e graduado em Engenharia

Elétrica. Foi o fundador e coordena desde 1996 a Divisão de

Tecnologias Tridimensionais do CTI Renato Archer. Sob a sua

supervisão esta divisão desenvolve aplicações e projetos de pesquisa

em cooperação com o setor produtivo e academia no Brasil e exterior.

É membro de vários comitês científicos, participante e palestrante

convidado nas conferências mais relevantes de impressão 3D.

[email protected]

CTI Renato Archer

O Centro de Tecnologia da Informação

Renato Archer é uma unidade de pesquisa

do Ministério da Ciência, Tecnologia e

Inovação (MCTI) que atua na pesquisa e no

desenvolvimento em tecnologia da

informação. A intensa interação com os

setores acadêmico, através de diversas

parcerias em pesquisa, e industrial, em

vários projetos de cooperação com

empresas, mantém o CTI no estado da arte

em seus principais focos de atuação, como a

área de componentes eletrônicos,

microeletrônica, sistemas, displays, software

e aplicações de TI, como robótica, softwares

de suporte à decisão e tecnologias 3D para

indústria e medicina.

131

Additive Manufacturing in the Product

Development

Abstract

Additive manufacturing (AM), also known as rapid prototyping or 3D printing, is nowadays considered a groundbreaking technology for the production of high added-value products with great investment from the developed countries. Additive manufacturing began as a way to produce higher quality prototypes more quickly, with minimum human intervention. However, what happened during the almost 30 years during which this technology has existed is that there was a great migration to various sectors of the industry, thanks to the evolution of the AM processes and their associated materials. It is already common to see highly customized products of AM being used in medical applications, or some not so customized products used in aircrafts. This paper has the purpose of summarizing the technological evolution of AM, pointing out its most important applications as well as some challenging applications in some sectors in which product development is a critical element. Additionally, this paper intends to present final use applications of this technology as well as what has been happening in terms of standardization in the area. Finally, some potential impacts predicted in the manner in which the products will be produced, as well as their logistics in the future, will be commented.

Keywords

Additive manufacturing; rapid prototyping; 3D printing; product development; product

customization.


132

1 Introduction

The 3D technologies are present in all domains, from the most simple product development,

optimization and production to the more complex products like in the aerospace industry. In a

broad definition 3D technologies can be classified into virtual and physical. The virtual ones

can provide computational models for many types of representation, simulation, optimization

and scientific visualization. The physical 3D technologies are responsible for transforming

virtual models into real objects. When using a layer-by-layer paradigm to build objects, it is

known as rapid prototyping and more recently called 3D printing by the non-specialized media.

The original name, rapid prototyping, was conceived in the sense that a prototype of a product

could be quickly and automatically created with any level of geometric complexity. A prototype

is the first of a series that is being used in many different industries to speed up new product

development cycle, lowering costs and increasing final quality.

Almost 30 years ago, a company called 3D Systems, today the biggest in this business,

launched in the USA the first commercial machine. It was the stereolithography apparatus

(SLA) invention of Charles Hull, co-founder of 3D Systems. In stereolithography equipment a

photopolymeric resin can be hardened by the incidence of a computer controlled ultraviolet

laser beam (Silva et al., 1999). Since then, a myriad of new technologies using many different

materials and processes became commercially available with many others in development or

at in research stage.

These new processes are everyday more present and focused in the real production every

day. In this context, the American Society for Testing and Materials – ASTM established in

2009 the Committee F42 on Additive Manufacturing Technologies to deal with this innovative

way of production. The Committee F42 scope is “the promotion of knowledge, stimulation of

research and implementation of technology through the development of standards for additive

manufacturing technologies” (http://www.astm.org/COMMITTEE/F42.htm). The very first

action of ASTM-F42 was to establish a standard terminology defining the official name of the

technology as Additive Manufacturing (AM). ASTM is integrating efforts with International

Organization for Standardization (ISO/TC 261) by means of a “Joint Plan for Additive

Manufacturing Standards Development”. The first joint plan session was hosted by ASTM in

June 2013. The plan is to integrate efforts in the area of test methods, processes, materials,

terminology and design in order to create international standards for the area. In this article

we will use, from now on, the term Additive Manufacturing or simply AM to be compliant with

the standards (ASTM, 2012) when referring to this technology.

The umbrella term AM encompasses a broad class of processes based on continuous

deposition of material, layer-by-layer, until a physical object is automatically built following

instructions from a computer with a virtual model designed in a Computer-Aided Design (CAD)

system (Figure 1). The material deposition can be achieved using highly specialized, high-end


133

machines, for serial production of metal, polymers and ceramics parts, or a composite

material, but can also be achieved with the new low-end so called “3D printers” that can be

acquired in many stores. The bigger companies on AM business are acquiring many of the

start-ups companies that produce this type of popular machines, closing the hardware and

software. Giants of the software area are becoming aware of importance of this type of popular

machines. Most recently, Microsoft announced 3D printing support for developers of Windows

8.1. Also, Autodesk (maker of AutoCAD) is supporting the development of specific software

to design human tissue and organs to be produced in the future in bioprinters, a variation of

AM for biological and medical use.

The AM is an evolution in course only compared to the personal computer (PC) revolution.

Today, a smartphone can have a processing power unimaginable 30 years ago with

insignificant costs and much higher processing capacity if compared to the former corporative

massive and expensive computers. Analogously, a 2D printer is today much more precise

costing thousands times less. The engineering development in electronics, software

development, control strategies, precise mechanics and communication (Internet) associated

to the new behavior in social networks, creates a fertile environment for AM evolution.

Figure 1: Physical object produced by a layer-by-layer deposition of material

The aim of this article is to highlight some of the AM processes, example of available

technologies and its applications and expected impact in high value manufacturing sectors.

Section 2 shows the categorization of Additive Manufacturing in classes of processes; Section

3 describes the AM main processes classifying them in one of the ASTM category; Section 4

brings a short discussion about the advantages of the AM in comparison with conventional

processes of production; Section 5 discuss materials available for AM and its challenges;

Section 6 highlights the AM as a convergence of the artisan production and mass production

as well as some applications that can corroborate this hypotheses. Section 7 highlights

potential impacts of the AM in global logistic. Finally, section 8 brings a general conclusion

about AM, including some expected impact it can cause in the future society.


134

2 Additive Manufacturing Categories

There are more than 30 commercial AM processes with economic importance. In order to

organize these processes in classes, in 2012, the ASTM International Committee F42 on AM

voted on a list of process category, names and definitions. The processes available nowadays

and in the future will be included in one of the following seven categories, according to ASTM

definition (Wohlers, 2012):

1 Material extrusion is an AM process in which the material is selectively dispensed

through a nozzle or orifice.

2 Material jetting is an AM process in which droplets of build materials are selectively

deposited.

3 Binder jetting is an AM process in which a liquid bonding agent is selectively deposited.

4 Sheet lamination is an AM process in which sheets of material are bonded to form an

object.

5 Vat photopolymerization is an AM process in which liquid photopolymer in a vat is

selectively cured by light-activated polymerization.

6 Powder bed fusion is an AM process in which thermal energy selectively fuses regions

of a powder bed.

7 Directed energy deposition is an AM process in which focused thermal energy is used to

fuse materials by melting as the material is being deposited.

There is a strong need for standardization and regulations for materials and testing of AM

products including all aspects of the AM technology. There is still a long way for AM to be

widely applied in areas like aerospace and healthcare. Therefore, processes reproducibility

and materials composition are necessary to produce high quality parts with required properties

and durability for the application.

3 Additive Manufacturing Available Processes

Among the dozens of commercially available AM processes, the most economically significant

are highlighted bellow and classified into one of the seven ASTM defined categories:

3.1 Stereolithography (SLA)

SLA was the first commercially available process in 1986. In this process a laser beam of

specific wavelength hits selectively the surface of a photopolymeric resin deposited in a vat.

The resin gets hard to form a layer. Then, a platform inside the vat is moved down and another

resin layer is photopolymerized. The process repeats until finishing the whole part. Support


135

structures are built with the same material and are broken down after taking out the part. It is

a process in the category of Vat photopolymerization according to ASTM. In the beginning,

there were few materials available but today there are many options. All of the available

materials, mainly polymers, try to imitate the properties of regular material for industry.

3.2 Fused Deposition Modeling (FDM)

FDM is a process where a filament of material and a filament of support material are

automatically deposited on a platform by means of a heated extrusion head. The process is

repeated layer-by-layer until the physical model is finished. By the end, the support material

is taken out by means of an ultra-sound bath. This process is in the category of Material

extrusion as defined by ASTM. This process produces accurate models in many different

thermoplastic materials, including special engineering thermoplastics.

3.3 Selective Laser Sintering (SLS)

SLS is a process where a laser beam transfers energy into a surface containing a thin layer

of pre-heated powder material. A computer automatically controls the movement of the laser

beam focus. The energy transferred by the laser beam fuses specific areas of the surface.

After fusing one layer, another one is deposited and again the laser fuses this layer that will

glue in the previous ones. This is repeated until the physical model is finished. The remaining

non-sintered powder is taken out after finishing. This process is located in the ASTM category

of Powder bed fusion. It produces strong models in many different materials and composites.

There are two other important commercial processes that can be classified in this category.

The first is the Direct Metal Laser Sintering (DMLS) that uses the same concepts with a fiber

laser to melts down materials with higher melting point like metals and alloys. The second

process is Electron Beam Melting (EBM) that instead of using a laser it utilizes an electron

beam to transfer energy to a metal or metallic alloy powder bed. SLS is highly suitable to

process composite and functionally graded materials with cellular structures.

3.4 Multi Jet Modeling (MJM)

MJM is an additive process where a print head containing hundreds of nozzles selectively

spreads in a surface a photopolymeric material that is cured and hardened by the incidence

of a specific wavelength – normally an ultraviolet laser or lamp. Another print head spreads

support material, normally in form of a gel that is also polymerized. It is a continuous process

where the platform is lowered down a tenth of a millimeter for each layer. Finally, the support

material is taken out with water jet. It is a process classified, according to ASTM, as Material

Jetting. The company Objet Geometries calls this process Polyjet. Some of the Objet´s

machines are able to combine different material within a single 3D printed model. This

company provides a myriad of material that mimics regular plastic materials.


136

3.5 3D Printing (3DP)

3DP is a process where a multinozzle print head selectively spreads a liquid binder in a

platform with a powder. The binder reacts with the powder to compose a layer while the

platform is moved down. The process repeats until the end of the part. There is no necessity

for support structures since the loose powder is responsible for stabilizing the part being built.

After finishing, the loose powder is taken out and the part is infiltrated with resin to increase

mechanical properties. ASTM classifies this process as Binder jetting. It is one of the most

popular today because of the low costs of acquisition and operation.

Figure 2 shows a diagram with the main commercially AM processing principle, materials

used, the process acronym and its schematic.

Figure 2: Schematic of the main AM processes, basic principle and type of materials that can be

processed

4 Additive Manufaturing versus Conventional Production

Beyond regular prototyping processes that were the first use of AM, this technology is

seriously entering the domain of production. Because its high flexibility AM presents many

pros and some cons when compared to conventional production processes. Among many we

can highlight the following:

Energy optimization – AM implies the use of energy only to transform the material only

enough. Conversely to conventional process, as metal milling process for example, in which

a material is transformed into a block of bulk material that is milled and the big amount of chips


137

has to be reprocessed (recycled) into a new block again and so on. According to the USA

Energy Department, AM can reduce energy costs in 50% and material costs in 90%.

Reduction in material waste – AM uses just enough materials to create pieces. However, this

is not completely true because sacrifice support structures sometimes are necessary and

some materials, in special polymers, it degrades with continuous use under heating.

Special tooling and speed – generally, additive manufacturing processes are slower than

conventional processes, but their speed is increasing in a very fast pace. There are some

advantages producing small and very complex parts when compared to conventional

processes. There is great flexibility to produce many different parts at the same time,

automatically, without the necessity of special tooling and fixtures. Today, the drawback of

additive manufacturing processes in comparison with conventional production processes is,

in general, the fast degradation of material properties (mainly polymers), stiffness, surface

finishing, costs and production time for big lots.

Design optimization – the use of additive manufacturing allows the designer to produce

complex shapes and moving parts without the constraints of conventional production.

Production of internal or hidden structures, conformal cooling channels, special passages for

cables, pockets for embedded sensors, actuators and optics, are also possible. The CAD

systems are not suitable for this task and new developments in CAD to represent complex

structures and diverse materials distribution from micro to macro sizes have to be developed.

AM will be useful and suitable to produce any structure for a simple product or for a complex

human anatomy, including organs mathematical representation. The latte will be responsible

for the development of a new class of specific CAD for healthcare as a BioCAD system (Silva

et al., 2012).

Figure 3 shows 10 principles of additive manufacturing that complement the previously

presented based on Lipson and Kurman (2013).

Figure 3: Ten principles of Additive Manufacturing, based on Lipson and Kurman (2013)


138

5 Materials Development for Additive Manufacturing

The ultimate goal of additive manufacturing processes is to produce physical models with

known and predictable properties. The commercial additive manufacturing systems

developers offer a huge amount of material choices but they are mostly specific for one type

of process and many times for a specific machine. There are developers that have almost 60

different materials for their processes. Each of their material offer specific capability to meet

specifications of form, fit, and function needs. Normally they defend their choices of proprietary

arrangements promising the best features you need but it is mainly a matter of Market. The

additive manufacturing materials are still very expensive proprietary solutions.

In some processes the development of a new material that behaviors “like” a well-established

material in industry is a challenge. Therefore you can see offers of the type “ABS-like” or

“Polyethylene-like” and many others that simulate the required features of a regular industrial

material but it is not exactly the same. Even when the suppliers specify their material as

“industrial grade” the final part produced using it does not have the same properties of the

regular injected or milled material, found in industry.

The development of materials that fit the AM process aiming certain application with a

minimum of waste and a maximum of functionality is paramount for the future of production.

Figure 4 depicts today’s possibilities of materials for additive manufacturing. These materials

will be shortly explained in the next paragraphs.

Figure 4: Materials for additive manufacturing

The polymers are the biggest class of material for AM. The polymers can be a thermoset or a

thermoplastic and be found in many different forms like liquid, semi-liquid, powder, filaments

and sheets, depending on the process to be utilized. The most important categories are the

epoxies, acrylate epoxies, Acrylonitrile butadiene styrene (ABS), Polycarbonate, Polyamide,


139

Polyphenylsulfone (PPSF), polyetherimide (PEI), etc. The most common processes cited in

this article that run polymers are SLA, Objet and SLS.

Metals and their alloys are the second biggest class of material for additive manufacturing.

Theoretically any metal material or alloy can be used in form of powder as far as the

temperature is enough to reach the metal specific melting point. On the other hand, there are

some intermediate processes that glue metal powder using a polymer and after the green part

is fabricated it is sintered and infiltrated by a lower melting point metal in an oven like in the

SLS process. Direct metal laser sintering and metal fusion are implemented by DMLS, SLM

and EBM proprietary processes. The most common metals and alloys are the various

Stainless Steel alloys, Cobalt Chrome, Titanium Ti6AI4V/ELI, Titanium Grade 2, Inconel,

Maraging Steel, Aluminum, and many others. The metal applications for added valued industry

and medical applications are growing very fast.

Ceramic is the smallest class of regular material for additive manufacturing but it is growing

significantly due to the possibility to modulate ceramic materials in complex geometry with

details not possible when using regular production processes. There are different ways to

process ceramic materials: a) Direct and indirect laser sintering of ceramic powders (SLS); b)

Movable print head selectively deposits a binder onto a platform (3DP); c) sheets of ceramics

can be laser cut, stacked and bonded using adhesives and heat (this process is known as

Laminated Object Manufacturing - LOM); d) ceramic particles in a semi-liquid or aqueous

suspension can be extruded into a filament (FDM) and e) ceramic particles can be suspended

into a photocurable liquid monomer, which can be selectively cured (SLA). The most usual

process is 3DP, where a liquid binder glues layers of material, and afterward it can be

processed in oven or infiltrated with resins to acquire mechanical resistance. Special sand can

be sintered by means of the DMLS process to create molds for metal casting. The production

of ceramic parts is very promising in industry due the possibility to produce very complex

shapes and modulate material properties. The ceramic area still lacks accurate process

development and applications. The materials mostly used are sand, gypsum, Zirconia, Silicon

Nitride, Alumina, etc.

Composite materials can be a mix or combination of the formers in any proportion and it is

very common to find materials like a polyamide matrix filled with ceramic or metal like glass,

aluminum or carbon fibers. There are also photocurable resins with ceramic particles,

especially for AM microstereolithography systems like Ormosil - organically modified silica and

Ormocer - organically modified ceramics.

New processes and materials are becoming available to produce final parts. A complex

combination of composite materials - with better properties than when they are isolated

materials; functionally graded material (FGM) - a class of advanced materials that varies the

properties along its dimension; and cellular structured materials – that can obtain an optimum

material distribution reducing weight and costs, are being researched and finding new

applications every day.


140

The recycling of plastic materials using AM can receive special attention from environmental

agencies and governments. There is a considerable amount of waste when post processing

for support removal or in powder based plastic material like SLS process that discards at least

30% of the material every batch of production, because of heating degradation of the polymer

powder. On the other hand, metallic materials produce little waste. Therefore, recycling is a

word that shall urgently enter AM dictionary.

Materials are responsible for a huge part of the AM providers’ incoming. The global AM market

reached $2.2 billion in 2012 with an increase of 28.6% for the last year, according to Wohlers

Associates Inc. a specialized consultancy in AM. It is still a small market (1,000 smaller than

conventional manufacturing in the USA) according to Lipson and Kurman (2013).

6 Uses and Applications of Additive Manufacturing

There are many well-known applications of AM. The initial purpose of the technology was to

create prototypes in a flexible and quicker way. With the evolution of the processes and

materials AM became a natural tool to solve some specific problems for small series direct

production (rapid manufacturing), tooling production, and more recently a powerful toll to

produce intricate and complex parts cost-effectively. AM is not yet a production process that

fulfills all the applications and industry requirements as dimensional, structural and surface

finishing. Parts produced in AM can be post-processed in conventional systems for a more

suitable and complete solution. Therefore, AM is not intended to substitute all conventional

production processes, mainly because of economical factors, but work integrated, at least in

the near future.

Although additive manufacturing is becoming an economically promising technology, it is now

almost impossible to expect that this technology can replace the mass production of parts or

components. It is still difficult to determine the cut-off volumes when comparing, for example,

plastic parts made on AM with injection molding production. On the other hand, AM flexibility

and increasing offer of materials can be a great solution for specific production of small series,

as well as for mass customization. Material and machines are expected to decrease price

occupying a more expressive market-share in the industrial goods sector.

Applications of AM are broad and present in areas like art, architecture, jewelry, consumer

products, entertainment, education, energy, electronics, paleontology, scientific visualization,

nanotechnology, automotive, aerospace, aeronautics, medical, dentistry, tooling, just to name

a few. Our mind is the limitation for AM application. The applications are driven today mainly

by the higher value-added goods and complexity of the solution that AM can help to solve.

Figure 5 shows the AM as a convergence of the benefits of mass production and artisan

production. In some extent AM can be close to the artisan solution or even the mass-

customization production but it is very difficult that in the future mass production will be

realized in AM due to the very specialized and optimized solutions like milling, plastic injection


141

and so on. Then, AM will play a special role in the industry but not substitute conventional

production as a whole.

The recent world crises, started in 2008, highlighted the manufacturing sector as one of the

most affected. It is the economical basis for industrialized countries corresponding to about

16% of the GDP and responsible for 30 to 55% of the jobs. The high value-added products

have an important participation in this percentage hiring a very specialized work force. In this

context, AM appears as one of the new enabling technologies for a series of advancements

and the basis for mass customization in the modern manufacturing (McKinsey, 2012).

Figure 5: Potential use of Additive Manufacturing as a convergent technology

In terms of the importance of the AM technology, every year Gartner Group assesses about

1900 information technologies in its report called “Hype Cycle for Emerging Technologies”. In

2012 AM (referred as 3D printing in Gartner’s report) is positioned together with its medical

variant “bioprinting” as one of the fastest growing for the next years (http://www.

forbes.com/sites/gartnergroup/2012/09/18/). In 2013, the same study reported AM in two

different points of the graphic “Customer 3D printing“ in the “peak of inflated expectations”,

probably because of the low-end printers being sold everywhere directly to customer and

“Enterprise 3D printing“ in the “slope of enlightenment” as a mature technology for industry

direct production but still far from its potential applications (http://www.gartner.com/newsroom/

id/2575515).

Today the most appealing areas for AM are the healthcare and aerospace industry. Medical

and dentistry demand highly customized solutions to fit a specific patient. Aerospace and

aeronautical industries involve highly complex product in small series. Both areas are strictly

regulated by national and international standards and agencies. These two applications are

more detailed bellow:


142

6.1 Healthcare

Today the application of AM in healthcare is one of the most promising using AM. The

fulfillment of needs and customization that the technology can reach is a good solution for a

specific patient’s needs. The most known use is the medical model (biomodel) for precise

surgical planning. A biomodel is a perfect replica of the patient produced based on

computerized tomography (CT) or magnetic resonance imaging (MRI) dataset of this patient.

The dataset is processed using specialized software that can generate a virtual 3D model in

a specific file format with the anatomical region of interest for AM production. It is possible to

export the processed files to CAD systems and generate customized prostheses for a patient.

During this process a simulation of the interaction implant-anatomy structure can be useful.

As an example, Figure 2 shows a complete cycle for a hemimandible production due a tumor

resection. Medical images from CT are acquired from the patient and a 3D model (STL file

format) of the mandible is produced using specific software (as InVesalius, an open source

software developed by CTI-Brazil) and translated into surfaces using the BioCAD approach

(Noritomi et al., 2011) (see Figure 6(a)). The missing part of the mandible is redesigned

considering biomechanical requirements. A mesh is then generated for simulation using Finite

Element Methods – Figure 6(b). The simulation involves bone and prosthesis made of

Chrome-Cobalt medical alloy with the DMLS process – (Figure 6(c)). The prosthesis is then

fabricated in the simulated material in AM (Figure 6(d)) (Delgado et al, 2012).

Figure 6: Implant production using additive manufacturing developed at CTI

Therefore, the above example shows that AM associated with other technologies can offer

solutions, for a temporary or permanent implant, with better functionality. So far, this

customized solution has not been cleared by the national agencies and its actual use depends

on approval from ethical committees.


143

The growing and aging of the population, mainly in BRICS countries, where the life expectancy

is growing in a fast pace can demand very specific solutions for healthcare. Brazil increased

life expectancy in 18 years in the last 50 years, reaching 74.6 years in 2012. It can be

translated into needs for high-performance orthopaedic implants for a longer and active

lifestyle. A growing number of companies with AM personalized solutions for healthcare,

already CE-certified and FDA-cleared, is blooming. One example is a company in Italy called

Lima that produces thousands of acetabular cups every year (http://www.lima.it/technology-

trabecular_titaniumtm-1.html). Beyond prostheses, orthopaedic tooling and instrumentation

are billionaire and very competitive markets involving a high quality control. AM is already in

use in cutting edge companies that produce thousands of these devices but there is plenty of

room for improvements, optimizations, customizations and time-to-market reduction. Dentistry

is already an area of consolidate applications and active developments using AM, like in

orthodontics, dental implants, drilling guides, dental molds, forensic, metal infrastructures as

bridges and crowns, just to name a few. The main AM providers are investing in materials and

specific processes for direct implant production in plastic, metal and ceramic, as a strategic

market.

In the multidisciplinary research domain the integration of AM and tissue engineering is

showing its preliminary results. The combination of biomaterials and AM can produce

structured replicas of an anatomy with strict controls of porosity that is colonized by patient

steam cells. After some time the material is reabsorbed by the organisms and a new tissue or

organ is created. This approach is called “scaffold tissue engineering” that is being improved

in a very effective way using AM (Pereira et al., 2012). More recently a new approach for

tissue engineering using AM has been investigated. This new approach considers the

automatic deposition of cells or macrotissues (called tissue spheroids) as an additive

processes to produce organs and complex tissues. This is called organ bioprinting (Rezende

et al, 2012) and promises to revolutionize the way tissue engineering is today, producing

organs ready for implantation. In the next decades, positive results of organ bioprinting can

overcome organ shortage and complex logistics of natural organs donation.

6.2 Aeronautic and Aerospace

Aeronautic and aerospace industry is one of the most powerful and a natural candidate for

AM using in the next years. Some applications are already in production and others in

research labs. This section will provide a short view of these applications.

NASA launched a program to include AM in its operations to produce parts for satellites and

space exploration vehicles, in space, in the near future. In this sense, they embraced a

program in cooperation with Washington State University intending to use a moon material

simulant with the same composition of ground moon surface dust to produce parts for a lunar

base by means of SLS process. More recently NASA at the Marshall Space Flight Center

tested two AM manufactured subscale rocket injectors in a hot-fire environment with


144

temperatures over 3,315oC. It was a monolithic complex part built in less than three weeks

with 60% costs reduction that stand-up for almost one minute with the same performance of

the injectors traditionally produced (http://www.nasa.gov/exploration/systems/sls/

3dprinting.html). NASA is also carrying out parabolic flight to improve AM equipment

robustness and operability to make them more suitable for space applications to deliver,

during long flights, on demand spare parts (http://www.space.com/16656-space-

manufacturing-3d-printing.html#sthash.5avSw7bg.dpuf).

According to Michael Idelchik, vice president for advanced technologies at GE Global

Research, in four decades from now on GE will be able to print a complete engine. If it will be

happen, only time will tell, but GE Aviation is investing hard and recently acquired Morris

Technology and Rapid Quality Manufacturing, two AM pioneers in aeronautics. GE Aviation

is also in a joint venture with the Snecma from France to incorporate AM produced combustion

systems in their newest jet engines (http://www.gereports.com/printing-jet-engines/).

Boeing is a pioneer and a long-term user of AM to produce parts for planes. Among those

parts there are air ducts for F/A-18 Hornet fighter aircraft produced from 10 years ago. The

ducts were originally intricate parts produced in aluminum. Now they can produce air ducts as

monolithic part in SLS process using plastic powder FAA cleared. The main gain is design

freedom, no assembling or welding necessary, reduced weight and consequently costs

reduction. Boeing produces its own high standard material with improvements in SLS process

stabilization and monitoring to produce high quality parts for its planes (Lions, 2012). One of

the recent patents claimed by Boeing is a new concept for SLS continuous linear production

for high production of parts (US Pat Appl. 20120067501). Another company in the AM parts

production for aircraft is CalRAM, a spin-off from Boeing, founded in 2005, that produces

plastic and metal certified components.

The AVIO Group in Italy designs, develops and manufactures aerospace propulsion

components and systems for both for civil and military aircrafts. To overcome traditional

investment casting and achieve outstanding properties, more complex and lighter

components, AVIO use Titanium aluminide in EBM machines and Stainless steel, cobalt,

nickel and aluminum alloys in DMLS process. AVIO also commercializes powder alloys

optimized for these processes (http://www.aviogroup.com).

6.3 Other AM Applications

Beyond the aerospace and aeronautical use, AM has been researched for complex

applications using precise new materials and processes. Size is becoming a very interesting

and important concern in AM. Today it is already possible to produce huge parts with four

meters wide. On the other hand, it is also possible to produce parts in the micro world with

details smaller than few microns, using processes like Two-Photon Polimerization (2PP)

(Ovsianikov and Chichkov, 2012). Nanoscribe GmbH, a spin-off of Karlsruhe Institute of


145

Technology (KIT), is producing what they claim to be the world's fastest AM equipment to

produce micro and nanostructures (sub-micrometer details) with Direct Laser Write

technology (http://www.nanoscribe.de). The microstructures can be potentially used in

healthcare (micro needles, stents, scaffolds for tissue engineering, etc.) 3D photonics,

biomimetics, microfluidics, mechanical microstructures and the polymer templates to create

metallic structures for electronics, among many others. Many universities, companies and

research institutes are investing in micro and nano AM technologies for advanced and diverse

applications.

Figure 7 shows in its right side the micro world of AM where small structures can be produced

like the micro airplane model produced in Karlsruhe Institute of Technology using its Direct

Laser Write technology and a functional monolithic micro turbines produced at Laser Zentrum

Hannover using a microstereolithography apparatus, both research centers in Germany. The

nano world is close. In the most right side of the picture can be seen a National Geographic

cover produced manipulating atoms at IBM Almadem research center. It is still a 2D (flat)

structure but opens up doors for a 3D structure. On the other hand, big structures like a 28

wingspan Unmanned Aerial Vehicle (UAV) was produced and tested by the American

Lockheed Martin, and whole buildings to be robotically produced as proposed by Professor

Behrokh Khoshnevis of Industrial & Systems Engineering in Southern University of California

using the process called Contour Crafting (http://www.contourcrafting.org/).

Figure 7: Range of sizes involved in Additive Manufacturing and its potential limits

The Institute of Photonics at the University of Eastern Finland and Dutch company LUXeXceL

are cooperating to develop what they call Printoptical Technology. It is claimed to be an AM

method for the production of optical quality components, such as lenses, without any need for


146

post-processing (http://www.uef.fi/en/home). At Printed Electronics 2013 – Germany, the

American company Optomec presented their process Aerosol Jet 3D to print a radio frequency

antenna and electrical power distribution for propellers and LED directly printed on the wings

of an unmanned air vehicle (UAV). The wings were fabricated using FDM process

(http://www.optomec.com/site/latest_news/news123). This is a potential for the consumers to

customize and demand online their electronic products in the future.

In the consumer market, it has appeared specialized companies offering customer-ordered

designs. You can choose hundreds of customizable and not customizable designs upload

your own or even sell it. The client can choose from plastic to ceramic or metal materials.

There are no equipments and tools for mass production and everything is produced in the

state-of-the-technology AM machines. That is the case of the Shapeways located in New York

(http://www.shapeways.com/), Thingverse (http://www.thingiverse.com/) and others.

Theoretically they can produce anything, from art to mechanical parts, jewelry, bags, glasses,

etc. It become clear a new way to produce any consumer products in a distributed high added

value way.

Figure 8 shows a Formula SAE competition car from the Mechanical Engineering Department

of the State University of Campinas (UNICAMP) that competed in 2008 with many tailored

parts made in CTI Renato Archer using the Selective Laser Sintering (SLS).

Figure 8: Formula SAE competition car using SLS tailored parts

Figure 9 shows a picture from the Brazilian National Institute of Space Research (INPE) that

is implementing the project Brazilian Decimetric Array (BDA) composed of a set of 26

antennas in a “T shape” configuration in the city of Cachoeira Paulista (Brazil) to monitor the

southern sky providing solar radio images for application to space weather forecasting. It is

innovative final use o AM with parts produced in polyamide using SLS process. This is the

first application of AM proposed by one of the 18 international partners. The parts are covering


147

the antenna and embedding all the electronics with forced ventilation from the underground to

keep temperature constant in all antennas.

Figure 9: (a) Array of antennas in a “T shape”; (b) design of antenna’s covers and housing for

electronics and (c) SLS parts mounted in one antenna

Another innovative use of AM in space was tested in 2007 when CTI Renato Archer sent a

chamber built to support a chemical reaction experiment at microgravity aboard the

International Space Station (ISS). The chamber was a complex experiment integrating

mechanical, electrical and fluidic systems with lightweight constraints defined by the space

program managers. Selective Laser Sintering (SLS) was the AM choice to build the chamber

in polyamide (Maia et al, 2008). The experiment is shown in Figure 10.

Figure 10: Left side: parts of the chamber to run experiment onboard of ISS. Right side: Brazilian

Astronaut Lieutenant Marcos Pontes during flight in ISS with the chemical reaction experiment in the

center of the photo


148

7 Additive Manufacturing and the Future of Logistics

Material development will make the difference for the use and expansion of AM into

production. In terms of logistics, machines will be transported once to their places, designs

can be transported by Internet but materials and its variety will be the most significant item in

terms of transportation, mainly if they will be produced in big industries, for the sake of

efficiency gains, instead of small ones locally distributed. Then, the global energy use and

greenhouse gas emissions will be involved in material displacement.

Therefore, a possible reduction in the logistics activities may be possible, since consumer

products can be locally produced in the site they are wanted reducing costs of transportation

and stocking. The products will flow directly from the producers to customers without retailers,

meaning fewer inventories. The task force will move from assembling lines to the new jobs in

small facilities. This new model will create opportunities for the development of new

management models and systems, new production models and controls and consumers

distribution even in more remote places.

According to United Network for Organ Sharing – UNOS (http://www.unos.org), a non-profit

organization in the USA, that manage a very complex net of information to optimize donation,

procurement and allocation of organs, a heart, for example, can be kept outside the body from

4-6 hours what makes a logistic, beyond biological and compatibility issues, very complex and

time dependent. The AM approach for organ bioprinting promises in the future to eliminate

waiting list and produce locally the necessary organs for transplantation.

Although AM is becoming an economically significant technology, it will be most improbable

that it can replace the mass production of parts or components, at least in the short to medium

term scenario. However, its flexibility and increasing offer of materials can be a great

opportunity for specific production or small series as well as mass customization. In the short-

term, the most appealing areas for the AM expansion are the aerospace industry, due to its

complexity with small series production, and healthcare, due its highly customized solutions

to fit a specific patient need. Assistive technologies for impaired people are experiencing a

fast growing but are still in development stage.

The implications of AM for logistics could be massive, both for the upstream supply chain and

for downstream customer order. AM will affect component suppliers, since the processes will

run in a single facility and retailers, because orders could come directly from consumers to

the near factory. It can potentially reduce warehousing, shipping and increase mass

customization (Manners-Bell and Lyon, 2012). American Department of Defense (DoD) is

seriously investing in AM to reduce costs of warehousing and transportation of spare parts to

the battlefield.


149

8 Conclusions

Additive Manufacturing has its root in the process for quickly developing products with the

basis in the first processes of rapid prototyping in the eighties. It is still consistently used in

the product development process but it is expected that AM can revolutionize the way

production and business are done today. AM is unquestionable in a fast-pace progress and

developed countries are strongly investing in the new production paradigm referred as the

“Third Industrial Revolution“ (http://www.economist.com/ node/21553017). This new paradigm

of production is expected to grow in the next decades. SMEs can potentially be the major

beneficiary of the new technology. The main aspects that can benefit SMEs are the local

production of highly customized products, facilitate logistics, and increase flexibility with lower

costs. The logistic industry can be strongly affected by AM but competitive companies in

business will find opportunity for the development of new applications and systems for

managing production, distribution, customer relationship, partner relationship, product-life

cycle, and many others. This new way to produce and trade will open up opportunities for

software developments and the way customer can interacts with producers.

Movement toward free software and hardware associated to the expiring patents of the main

commercial processes created the “customer additive manufacturing” market with affordable

and popular solutions bringing 3D printers to homes in the last 5 years. Nowadays there is an

explosion of small companies offering very cheap DIY kits or mounted low-end machines for

makers or hobbyists. On the other hand, new horizons on the open-design direction with

people exchanging experiences and design solutions via Internet, like the today’s open-source

software and hardware, will dramatically affect industry and its intellectual property policy in

the way it is done today.

The main current industrial applications are the markets for customized, short-run items, such

as dental products, hearing aids, and jewelry, but some applications will highly impact lifestyle

like the healthcare and aerospace industry. In order to fulfill the challenging requirements of

economical and advanced applications AM has to grow in directions like: printing large

volumes and large objects economically; expand the range of printable materials; reduce the

costs of printable materials with higher durability and quality; monitoring of quicker processes

to guarantee reproducibility and surface finishing; and printing of multimaterials like those for

electronics. Legal responsibility will be a great concern demanding regulations and standards

for processes and materials testing.

9 References

ASTM (2012). http://www.astm.org/Committee/F42.htm

Delgado, J.S., C.A.R. Laureti, A.A. Camilo, J.V.L. Silva, L. Serenó, J. Ciurana (2012).

Mandible Reconstruction Using an Additive Manufacturing Technology In:


150

Proceedings of the 1st International Conference on Design and Processes for

Medical Devices PROMED. (Ceretti (Ed)), 275-278. Neos Edizione srl, Rivoli.

Lions, B. (2012). Additive Manufacturing in Aerospace. Examples and Research Outlook. In:

The Linking Engineering and Society Bridge. (Latanision R.M. (Ed)). 13-19,

National Academy of Engineering.

Lipson, H. and M. Kurman (2013). Fabricated: The New World of 3D Printing. Wiley.

Manners-Bell, J. and K. Lyon (2012). The implications of 3D printing for the global logistic

industry. Transport Intelligence. 1-5.

McKinsey (2012). McKinsey Global Institute Report Manufacturing the future: The next era of

global growth and innovation.

Maia et al (2008) Maia, I.A., M.F. Oliveira, P.Y. Noritomi, J.V.L. Silva. Application of Rapid

Manufacturing to build artifacts for using in microgravity environment. An

International Space Station case. Virtual and Rapid Manufacturing. (Bártolo (Ed)),

559-562, Taylor and Francis Group, London.

Noritomi, P.Y., T.A. Xavier, J.V.L. Silva (2011). A comparison between BioCAD and some

known methods for finite element model generation. Innovative Developments in

Virtual and Physical Prototyping. (Bártolo (Ed)), Taylor and Francis Group,

London, 685-690.

Ovsianikov, A. and B. Chichkov (2012). Three-Dimensional Microfabrication by Two-Photon

Polymerization Technique Computer-Aided Tissue Engineering, Methods in

Molecular Biology, 868, 311-325.

Pereira T.F., M.A.C. Silva, M.F. Oliveira, I.A. Maia, J.V.L. Silva, M.F. Costa, R.M.S. M. Thiré

(2012). Effect of process parameters on the properties of selective laser sintered

Poly(3-hydroxybutyrate) scaffolds for bone tissue engineering. Virtual and Physical

Prototyping, 7, 275-285.

Rezende, R.A., F.D.A.S. Pereira, V. Kasyanov, A. Ovsianikov, J. Torgensen, P. Gruber, J.

Stampfl, K. Brakke, J.A. Nogueira, V. Mironov, J.V.L. Silva (2012). Design,

physical prototyping and initial characterisation of Lockyballs. Virtual and Phys.

Prototyping,7, 287-301.

Silva, J.V.L., C.E. Saura, M. Bergerman, M.C. Yamanaka (1999). Rapid Prototyping -

Concept, Applications, and Potential Utilization in Brazil. In: 15th Intern.

Conference on CAD/CAM, Robotics and Factories of the Future.

Silva, J.V.L., T.A.P. Martins, P.Y. Noritomi (2012). Scaffold Informatics and Biomimetic

Design: Three-Dimensional Medical Reconstruction. Methods in Molecular Biology.

1edition, Humana Press, 91-109.


151

Wohlers, T. (2012) Wohlers Report: Additive Manufacturing and 3D Printing State of

Industry, Wohlers Associates.


152


153


154


155


156


157


158


159


160


161


162


163


164


165


166


167


168


169


170


171

172

MSc. Eng. Erwin Karl Franieck

Erwin Franieck was born in 1961 in São Paulo. Mechanical engineer

graduated at the Universidade Estadual de Campinas in 1985. In 1986

started his professional career as product development engineer for

the “Gasoline Systems-GS” Business unit at Robert Bosch Ltda,

where he was responsible for the implementation of laboratories for

functional analysis and electronic injection system component’s

calibrations. In 1991 he was transferred to the systemists unit,

becoming responsible for the mechatronic system integration,

managing the fuel feeding, ignition and torque control of the motor for

the vehicles and motors from 1992 to 1999. During this time, he

worked in several pioneer innovation projects, among them was the

Flex Fuel injection system. In 2001 he went on to participate at the

company’s headquarters, in Stuttgart, on the structuring of the

globalized business units, strengthening the connections between

these units among continents. In 2003, after returning to Brazil,

became responsible for the department of product development’s

engineering and expanded the fuel’s pump market share from 25% in

2003 to over 90% in 2008 by making use of the quick industrialization

of the flex fuel systems. In 2010 he was invited to manage an ISEC

(International Simultaneous Engineering Center) project of a new

technology responsible for coordinating both the project team as well

as the manufacturing team until the starting of production in 3 large

OEMs. Mr. Franieck later on returned to Brazil in 2013 where he

became responsible for the engineering directorate of the GS unit.

Once becoming director of the engineering unit, Erwin coordinated not

only the product development team in Brazil, but also the development

laboratories, accumulating during these 28 year of experience dozens

of patents. Mr. Franieck is currently focusing on the dissemination of

the Design for Six Sigma’s methodology, flexible management of

projects and integration of the value chain of new products which are

based on 3D files.

[email protected]

Robert Bosch Ltda.

The Business Unit Gasoline Systems (GS-

LA) is the greatest business unit of Robert

Bosch Brazil, located in Campinas,

completing 60 years of Brazil in 2014. With

revenue of R$900 Mi, GS-LA has an

engineering team of approximately 200

engineers working on product development

and research for 30 years in Brazil. As one of

the business units with greatest potential for

innovation in automotive segment, acting in

the development of products for admission

system, exhaust, supply and dosage of fuel

besides sensors and the electronic

processors units; GS regularly presents new

technological launches with the launches of

each OEM. The technological bases for

developments in Brazil present 3 basic

pillars: The Central Unit of “Corporate

Research - CR” in Germany, where basic

researches are made, the Centers of

Competencies (CoC) for certain

technologies, as fuels, polymeric materials,

metals, etc, situated in many locations (GS-

LA is CoC for alternative renewable fuels),

and the research institutes in Brazil.

173

Bosch Engineering System – A Robust Design

Process and 3D Model Applied in the

Complete Product Development Chain

Abstract

Bosch Engineering System (BES) is an integrated system for product development, based

on the deep knowledge of the product and on the cause-effect relationship between

product requirements and their functional characteristics, promoting the pursuit of

competencies in the centers of excellence in each technology and usage of development

tools and product simulation with full domain of them. Besides taking care of the involved

engineers’ competencies, the BES evokes the management to participate by managing

the team based on “content driven leadership”. We are going to approach an example of

Robust Design applied at RBLA. With one robust design available and also a clear idea of

the involved manufacturing processes we pursuit to automate the generation of production

data. Based on a 3D file elaborated in an adequate way including tolerances for the

manufacturing process in 3D, (i.e.: injected parts must foresee the split line and the draft

angle, etc.) through usage of PLM tools and simulations (i.e.: Mold-flow) we obtain one

geometry ready to initiate the tooling design. With this data, the responsible to develop the

tooling can automatically correct the contraction of the injected part obtaining the mold

geometry to be machined, transferred in an automatic way to the machines which will

produce the tooling. Once the tooling is ready, the try-out is started, injecting the first parts.

With tridimensional measurements through 3d scanning, it is possible to have a cloud of

the measured points and then comparing them with the product 3D file, automatically

evaluating if the samples dimensions are correct according to the tolerances. With the data

from the measurement, it is possible to use them to automate the correction of the tooling,

without been necessary to measure each dimension separately. This is a huge advantage

in time and recursions in the projects. Finally after the tooling correction, new

measurement and approval with less recurrence, the own 3D file can be used to document

into the quality control of the involved parts, registering in 3D e visualizing these files in

3D viewers already available. This is already a reality which brings competitiveness,

velocity and elimination of fails and recursions in the industrialization of the product.

Keywords

BES; Design for Six Sigma; DFSS; DRBFM; DOE; Simulation; PLM; 3D Scanner; 3D

dimension; Lean Development Process.


174

1 Robust Design Process

“A development process for designing robust products” is a mandatory discipline for all

enterprises in the automotive branch or others related to the products in the high value

assemblies systems. The challenges involved to keep the development process of new

products as a success factor in our days, are increasing quickly. Some causes of that can be

attributed due reduced product life cycle, increased warranty period to 5 years, complex

market regulation for environments (gas emissions, fuel consumption, CO2 limits, self

diagnosis, etc.), integration of several mechatronics systems, expand bio fuels technology,

electrification, among others.

There are mainly two ways to innovate in a market: one is to make evolutionary innovations

according to market´s demands and the other is to come up with a disruptive innovation, that

is, with completely new features and ideas. Tesla, a new car maker brand, bases their

products on a disruptive innovation by producing electric & connected cars with better

aerodynamics, user friendly interfaces with Internet etc., while also making use of evolutionary

innovations. This combination has granted Tesla a significant market share on the U.S.

Looking at Tesla as an example, an important question companies need to make and be able

to answer is: “How do I keep myself competitive on the market?”.

There is no pretension to present in this few pages a solution for this challenge, but to give a

summarized view from a big player on this market, Robert Bosch GmbH.

The structuring of the product’s development process, foresees an integration of areas and

functions, which are considered primordial for the execution and completion of the project (as

shown in figure 1). With this figure we can set the structure in order to attend the local demands

in a flexible way, while making sure not to lose the focus on experience and global know how.

An integration of each individual’s characteristics and team work dimensions have to be clear

structured and aligned in very competent portions to consider adequately the demands of

each type of project.

The Figure 1 shows the main dimensions of system development, integrating matricial

management teams to the product development process.

This sub-structure on Figure 1, called BES (Bosch Engineering System) can be connected

with other sub-structures like: Bosch Production system (BPS), Bosch Sales & Marketing

System (BSS) & Bosch Human Resources System (BHS), implemented worldwide in an

integrated business System.

The combination of relevant knowledge (KM) in competence centers integrating a global

Network and an idea generation process (IM), will stimulate the generation of ideas that are


175

rated and incorporated in each location, according to the targets of each regional business

unit.

Figure 1: Product Engineering System

Systematic design development with passion for engineering and identification with Company

Vision: “Technology for Life” are considered as essential success factors. To reinforce the

effectiveness of this affirmation, a competence management, which warranties the knowledge

and resources involved in each system are under the domain of the participants of these

project teams.

Some basic rules for the development team are to be considered in the early phase of work,

in order to investigate the alternatives and solutions for the requirements:

1 Establish a frame work and structure for content-driven engineering and content-driven

leadership at management teams.

2 Coordinate all engineer methods/ applications needs, (QFD, DFSS, DRBFM, DRBTR,

Powerful 3D Interface for CAD/CAE/CAM, etc).

3 Stimulate engineer work through focus on technical models with a deep understanding

of cause-effect relationships before going to the final design.

4 Stimulate the team to bring innovative solutions mainly in evolutionary form, managing

ideas that are not applicable at this moment, but will be registered in an Ideation system

for future projects.

The expected effects after application of the aforementioned set of rules are well known, and

can be summarized in Figure 2, while bringing a deeper knowledge of the functions of “the

Design Elements” (DE). One of the most important consequences is the effective reduction of

recursions during release phase, increasing the first pass yield quota, avoiding problems after

the start of production (SOP).


176

Figure 2: Start the development effort in early phase

Experience shows us that, an empirical approach at the start of designing new products is one

of the most common failures in the landscape of field claims. “Show me how your project starts

and I can tell you how it will finish.” Gero Lomnitz.

The analysis of field incidents shows that the most important reasons of field crisis are related

to:

1 Poor evaluation on requirements for each application.

2 Inappropriate test methods tools.

3 Poor evaluation of changes in the product/process.

4 Poor risk evaluation on Management from business units.

5 No usage of simulation tools.

The main efforts in the early phase of projects require a deeply understanding of product and

system requirements, which unfolds each and every product’s functions showing its main

components / sub-systems and helps in the thorough search and finding of the Modules and

Design elements in order to ensure the best solution for the achievement of the requirements.

Continuing with the product’s development process, as can seen on Figure 3, based on the

Design Elements, a cooperation process between the “Engineering System” and “Production

System” is started, working to find the better relation among Function Design Process

Robustness. A strong project management leader is important in order to ensure the

smooth cooperation among the involved work teams and the structures, considering the time

and budget constraint.


177

Figure 3: Management of Product Requirement

2 Robust Simulations

“The robust simulation of reality due to the virtual technology” is an essential concept in

the product development process today. This reality can be defined as the physical and

chemical phenomena that a product is subjected during operation or during the manufacturing

process. In the current scenario of automotive engineering, the robust simulation consists in

applying computational numerical methods (CAE: Computer-aided engineering) combined

with mathematical optimization methods.

Figure 4: Definition of a Simulation Model to represent the products functions


178

The utilization of a CAE method requires the creation of a parameterized mathematical model

able to represent the functions of the product satisfactorily, as shown in Figure 4. Generally,

a product has different functions in different physical domains. In this case, systems composed

of different models should be created in order to represent all of the products functions, as

illustrated in Figure 5.

Figure 5: System Simulation – A composition of different models

A system or model has normally a lot of parameters or variables that influence the

performance of the product. The simulation of all possible combinations of these variables is

unfeasible. Therefore, it is customary to use mathematical optimization methods, such as DOE

(Design of Experiences) or EA (Evolutionary Algorithms). The efficiency of these methods can

be defined as the ability to determine an optimal configuration of parameters with the lowest

number of simulations process.

The GAs (genetic algorithms) can be cited as an example of robust simulation [1]. This is a

method of EA that belongs to the field of artificial intelligence. In this case, the evolution of

generations of genetically different individuals is used to represent the search for a great

product, as shown in Figure 6.

Each product is represented as an individual, where a group of individuals is defined as a

generation. The new individuals of a new generation are created through a crossover between

individuals (combination of parameter values between different products configurations) and

/ or mutation of individuals (random modification of parameter values of a product). The

individual with higher fitness (product with the best performance) has a statistically greater

chance of passing on their genes through the generations. Thus, the optimal product is

determined to be the best individual of the last generation.


179

Figure 6: Genetic Algorithms – An Example of a robust simulation method

Currently, the robust simulation is increasingly present in product development processes,

mainly due to increased computing performance. This type of simulation favors the

understanding of the functions of the product and helps in taking decision regarding the

geometrical design or choice of materials.

In manufacturing processes, such as injection molding, it is also possible to choose the best

parameters of injection that generate lower costs and lower manufacturing time. Figure 7

shows an example of a performance comparison between different injections molding process

configuration.

Figure 7: Moldflow 3D Warpage Analysis Results with different injection points configurations


180

Besides seeking the great product, the robust simulation aims to reduce the cost and time to

develop a new product. The development engineer can interpret this simulation process like

a “robust virtual test bench” where the repetition of a "virtual experiment" is quick and

cheap.

3 Robust Design Based on Design for Sixt Sigma (DFSS)

“A DFSS approach demands deep understanding of requirements at Design Element” Here

we will make a brief description - based on a textbook example - about the basic elements,

for which deep knowledge of the interrelation among them for each Design Element is of

crucial importance.

We have 3 basic Elements that are interrelated:

1 Load - What is acting on the product

Sum of the mechanically, chemically, thermally and

electromagnetically induced loads externally applied

on the product.

2 Stress – What the Design Element

(DE) senses

Local effect of the load within the

design element* with respect to the

considered damage / failure

mechanism.

E.g. mechanical stress, induced

voltage, temperature distribution, mass transfer during a chemical reaction.

3 Strength – What the DE can endure

Maximum stress endurable by a design

element for a specified amount of stress and

for the damage-/failure mechanism under

consideration.

Calculation of stress from

measured load

Conversion of stress into amount of stress

Amount of stress

Lifetime curve

Sterength for amount of stress and damage/failure mechanism

Lifetime under this amount of stress

Str

ess

[MP

a]

Lifetime (number of load cycles)


181

3.1 Damage Mechanism – How the DE Fails

A damage mechanism includes all processes in the design element considering the duration

and intensity of the stress combination leading to a gradual change of its properties.

All damage mechanisms are classified in the categories of Cracking, Aging, Corrosion,

Biological material damage and Wear. This describes the processes resulting in failure (=

irreversible loss of function).

The Stress-Strength Interference Model of Reliability. The Robustness in this case, is

represented as being the distance between the statistical distribution of the Stress-curve in

relation to the Strength-curve. Thus, a design element fails if, at the failure location (i.e.

locally):

A stress occurs which exceeds the product’s strength,

There is poor strength at that location.

Based on this understanding, for more complex real life situations, it is important to assess

the solutions by making use of tools of best practice, lessons learned, etc, while also looking

for alternatives and using decision tools in order to pick the best solution. After the alternative

is chosen, it is necessary to define the parameters of the Design Elements.

For the elaboration of the cause and effect model, the knowledge about the interaction

between the parameters regarding the Design elements is essential. In the model presented

on the diagram in Figure 8, we can see the triad: Simulation, Calculation and Experiment

provide results that enable the adjustment und comprehension about all phenomenons

involved in the Design Elements.

The development activities will work to find the better solution to increase the strength of the

product/system in order to support the new requirement.

Calculation / Theory Study

Simulation


182

Experiment

Calculation / Theory Study

Figure 8: Diagram Cause-effect Relationship

4 Real Case: Improve Strength to Support New Requirement

Let us consider a typical application in the Brazilian market: we have a specification for the

fuel, nevertheless, we often find cases that do not meet this specification. In this case, a

reassessment of the real facts brings a revision of the requisites for the Fuel Pump product as

well as the emergence of new technologies, as for example a smart generator.

Product / Components study and functions mapped: Change of requirements for wearing on

electrical commutation for a Fuel Pump in ethanol fuel. (New requirements: Intelligent

alternator with variable voltage and bio fuel with higher conductivity). Electrical Arcing on

commutation area = f (residual energy; dielectric resistance; voltage) [2, 3].


183

1 Residual energy = f(commutation angle)

2 Dielectric Resistance = f(fuel conductivity (µS/m))

3 Energy = f (voltage)

Simulation: The Calculation/Theory output pointed the voltage and commutation angle

as the most relevant influence on spark energy under the new requirement (increase

stress).

Residual energy = f (commutation angle, Voltage)


184

Experiment (Design of Experiment): The definition of the experiment that reproduces the

failure mode is a fundamental step for the assessment of the actions and verification of

their respective effectiveness. For this, as seen in Figure 9, the awareness of the failure

mode and the influence of the new requisites are essential for defining all the details of

the experiment.

Figure 9: DoE: Design of Experiment: Test definition to check the robustness of The Design Element

Targets for the DoE:

1 Implementation of a failure reproduction method

2 Control of stress during the tests.

3 Finding the relationship of the design elements and strength curve.


185

Figure 10: Example of Results for influeny of 3 Parameters (Fuel Conductivity, Voltage, commutation

angle) on Life time during the DoE

Conclusion: The deep understanding of the phenomena, checked with appropriated

simulation tool and confirmed during DoE tests, gives the development team a basis for a

correct decision of proper design element parameter (commutation design), minimizing the

risks of recursions during release phases or after SOP.

5 CAD-CAE-CAM Tooling Supporting Design 3D

After defining the Proper Design Element parameters and limits necessary to achieve the

requirements, the complete design element starts to be fixed. This is made by thinking on the

complete product development cycle since the concept phase, passing through discussion

with production process and assurance of appropriated quality level. An integration of several

internal and external areas of product and process development (that in most cases are not

under the same business unit) is required.

Throughout the different phases of a Project, the transferred documents and data (drawings

and product specifications) between the countless departments (purchase, process planning,

manufacture, quality control and documentation) is most usually made via 2D documents.

Nevertheless, a re-edition of these drawings and specifications is needed in each transferring

phase of the documents in order to attend the necessary needs of each department. The issue

arises due to the fact that only some primordial characteristics concerning the manufacturing

process like plastic Injection Molding, cold forged metals, metal molding, etc. are considered

in the first 2D design. As a result of that, due to deeper discussions concerning the

manufacturing process, many new details demanding attention arise on subsequent phases

of the project that were previously not accounted for. In order to solve this detail issues, it is

usually necessary to review the initial project and sometimes even, in some cases, to modify


186

the project itself. This generates extra work, delays and resources wastage in an already

advanced stage of the project.

As a result of that, this leaves the engineer on one side, with the time constraint to release the

drawing to start the tooling preparation, and on the other side, the tooling maker, that will have

a lot of problems to achieve the requirements. Others involved are the quality control

department and the production department that must find a way to check if the parts are in

accordance to the drawings and how the process control must be managed to warranty the

product requirements.

A lot of recursions are expected during the development cycle, basically because the product

drawing is not focused on process, but on the functional elements.

CAD/CAE/CAM software have powerful solutions to check the 3D drawing according to the

process. If the product engineer not only has to be responsible for the functional requirements,

but also for the production requirements of a given product, he will have a lot of software

modules in hands that can support him on how to correct the design in an early phase of the

project according to the process demands. The computer aided 3D design can be used to

define not only the complete geometry of produced parts, but also the dimensioning including

dimensional tolerances, geometrical tolerances, drawing notes, specifications, etc. At the NX

CAD/CAE/CAM software, used at Bosch there is a PMI module (Product Manufacturing

Information) that follows several norms for 3D Digital product definition, which allows to include

in the 3D Design all PM information linked to the related dimension or surface.

Considering the 3D data with PMI as a full documented file, that can be exchanged in an open

3D format (.JT extension) and used in all kind of CAD software, it is permitted to exchange

the 3D Design and PMI among all involved areas, substituting with higher efficiency the 2D

drawings and specifications.

Each company has to establish some rules for using this kind of powerful resource, and keep

some internal standards, to speed up the exchange of information among the departments

and product life cycle.

A management of this UX (User experience) software is enabled through Teamcenter

Engineering (TCEng.) application. Among many function, it is used to create a workflow in the

3D drawing release process, which can include a lot of checks about the maturity of the

drawing. A powerful checkmate named HD3D has a customization for plastic parts, named

DFMA, that will require the trend line for the tooling and check automatically the minimal

radius, angles, wall thickness, etc.; controlling the release of the 3D-drawings for the

production system.

Another SW tool to guide the 3D design work for plastic parts, named Mold Part Validation

(MPV) helps the engineer to apply commands appropriated for plastic parts, simplifying

changes, accelerating the development time, with a proper overview about the process


187

necessities. With the design of the features in 3D, it is possible to define combined surfaces

as a complex format, that is, as one group of dimensions with the similar or same tolerance

and surface requirement. These can substitute a lot of non critical dimensions per the 3D

complex format, without a lot of dimensions.

The new equipment for measurement via White light or 3D laser measurement, delivers a 3D

cloud of points integrated in a surface that can be compared with theoretical 3D data, and

reduces the necessity measurement interpretation and individual dimensional measurement.

An example of application of 3D drawing with PMI and 3D Measurement in Figure 11 shows

the potential in time improvement for projects and products to reach market maturity.

Figure 11: 3D Scanner Blue light with precision of 0,030 mm

The usage of the 3D GD&T and the 3D scanner, in comparison with Standard 3D Optical

measurement results in a significant reduction of measurement time for complex parts, as

presented in the Figure 9.

By using PMI, Scanner and CMM, we are able to reduce the quantity of dimension by ~68%

(from 248 to 78) and the entire missing details can be taken from the model, see the note:

The accordingly usage of 3D tools in all phases of development for intermediate complexity

plastic parts for example, will significantly improve the project’s Time to Market. A first increase

in time, in order to do the design freeze including the demands of processes in an early phase,

will be more than compensated by reducing the recursions, and improving the timing for each

round of tooling try-out.


188

Figure 12: Reduction of measurement time using 3D Programming and #D Scanner


189

Conclusion: Innovation is a key success of factor on current product development process on

automotive industry. Pressure on time and costs combined with increasing level of complexity

of products and its interaction led to constant adoption of so called state-of-art methods and

tools to assure robust design and support organization all over development chain through

manufacturing. Engineers count even more with established IT-tools to obtain a virtual 3D

model prototype which can foresee and predict potential product improvements and based on

that perform design changes accordingly. In addition the information flow is streamlined once

the same 3D model is used throughout organization from concept till inspection. This

systematic approach results in effective reduction of recursions during release phase,

minimizing costs and increasing product quality.

6 References

[1] Michalewicz, Z.

Genetic algorithms + data STRUCTURES = evolution programs, 3rd ed.

Springer, New York, 1996.

[2] Sawa, K.; Wei, C. C.; Ueno, T.

Erosion by Arc Discharge at Carbon Contacts in Various Automotive Fuels, Holm

Conf. on Electrical Contacts, IEEE 59th, pp.1-6, 2013.

[3] Sawa, K.

Arc Discharge and Contact Reliability in Switching and Commutating Contacts,

Electrical Contacts, Proc. 51st IEEE Holm Conf., pp.10-21, 2005

190

MSc. Eng Waldir Gomes Gonçalves

Waldir Gomes Gonçalves graduated in Naval Engineering, and has a

Master degree in Aeronautical engineering with specialization in

composite structures. Working in the product development

engineering for more than 25 years, has participated in several

Embraer product developments, and headed many engineering

areas: Structures and Materials, CAE/CAD and PLM, Product Support

engineering, the Commercial Aviation engineering, the Embraer

Defense and Security Engineering and currently the KC-390 Product

Development Engineering.

[email protected]

Embraer S.A.

Embraer is one of the world’s leading aircraft

manufacturers, a position achieved through

the commitment to full customer satisfaction.

Throughout its 45-year history, Embraer has

been involved in all aspects of aviation. In

Addition to design, development,

manufacturing, sales and technical support

for commercial, agricultural and executive

aviation, Embraer also offers integrated

solutions for defense and security and

services.

191

Processo de Desenvolvimento de Produtos

Aeronáuticos

Resumo

O Processo de Desenvolvimento de Produtos (PDP) Aeronáutico se destaca como

vantagem competitiva. Para desenvolver um sistema sócio técnico de alta performance

são necessários vencer diversos desafios associados à aplicação eficaz dos princípios do

Lean Development. Por meio de uma gestão de processos, gestão integrada de

programas, desenvolvimento enxuto do produto e da gestão do processo produtivo a

empresa desenvolveu um PDP Aeronáutico com práticas integradas. A aplicação destas

práticas é demonstrada no estudo de caso do Desenvolvimento do KC-390. Este é o

desafio que a EMBRAER S.A. vem trilhando nos últimos sete anos.

Palavras chave

Desenvolvimento Enxuto do Produto; Sistema Sócio Técnico; Lean Development.


192


193


194


195


196


197


198


199


200


201

202

Mauro Conceição

Mauro da Conceição é Tecnólogo em Processos de Produção e

Gestor Global de Kow-How Management da Magneti Marelli Cofap,

com MBA em Conhecimento, Tecnologia e Inovação pela FIA –

Fundação Instituto de Administração e Especialização em

Gerenciamento de Projetos.Funcionário da Marelli há 24 anos,

atualmente é responsável pela administração e suporte aos

Processos de Gerenciamento de Ciclo de Vida do Produto no Brasil,

Europa, Estados Unidos e Joint-Ventures, incluindo administração e

suporte a todos os softwares de Engenharia, incluindo as ferramentas

de CAD e implantação de melhores práticas em

modelamento.Também foi coordenador do Projeto SABER na

Diretoria de P&D, que identificou e mapeou os processos críticos para

o desenvolvimento de Produto, gerando um diagnóstico do nível de

maturidade em relação às práticas de Gestão do Conhecimento.

[email protected]

Magneti Marelli Cofap

Fundada em 1951, a antiga Cofap Cia

Fabricadora de Peças, foi comprada em

1997 pelo Grupo Magneti Marelli.Hoje, a

Magneti Marelli Cofap é uma das divisões da

Magneti Marelli, multinacional italiana da

Grupo Fiat, que além da Linha de negócios

de amortecedores, possui as divisões de

Iluminação, Powertrain, Eletrônica,

Exaustão, Plásticos e AfterMarket, somando

no total aproximadamente 38000

colaboradores ao redor do mundo.No Brasil,

é líder no fornecimento de amortecedores

automotivos para as grandes montadores,

com mais de 70% de Market share.

Atualmente possui centros técnicos de

aplicação no Brasil, Estados e Itália, além

das unidades produtivas no Brasil (Lavras e

Mauá), Polônia (Bielsko), Estados Unidos

(Pulaski) e Joint-Ventures na Índia e China..

203

PLM na Magneti Marelli Cofap:

Compartilhando um Caminho, Dificuldades e

Desafios na Implantação Globalizada

Resumo

Em 2004 foi iniciada a implantação do Projeto PLM (Product Lifecycle Management) na

Magneti Marelli Cofap, unidade de Amortecedores do Grupo Magneti Marelli. O que se

pretende abordar neste caso de implantação é além de mostrar entregáveis e detalhes

técnicos da funcionalidade de um software, também trazer experiência sobre o ambiente

de implantação, o impacto das mudanças em uma organização que deixou de ser

nacional a partir da aquisição para a Magneti Marelli e as lições do aprendizado e boas

práticas que podem ser adotadas em projetos cuja configuração é de alto risco, pelo

escopo amplo, pela longa duração e diferentes culturas envolvidas.

Palavras chave

PLM; Implantação; Integração CAD-PLM; Workflow


204

1 A Tecnologia em Evolução

Amortecedores são pouco conhecidos, não possuem design atraente, ficam ocultos embaixo

do automóvel, porém são itens de segurança, com comportamento fluidodinâmico complexo,

cuja aplicação pode ser em veículos pesados, automóveis de passeio até o segmento

Premium para Maseratti, Ferrari com sistemas de amortecimento variável.

É necessário entender esta dimensão, porque a complexidade e variabilidade de portfólio

acabam por se tornam também variáveis que iram demandar um controle cada vez mais

confiável em relação gestão de mudança e busca de informação.

2 Histórico

Tudo começou como Cofap, quando na década de 50 Abraham

Kasinsky fundou a Cia. Fabricadora de Peças, que se tornou a

maior fabricantes de autopeças do Brasil, mas durante o processo

de abertura de mercado iniciada na era Collor e os problemas de

sucessão foi adquirida pela Magneti Marelli em 1997,

multinacional italiana que buscava integrar ao grupo a expertise

na fabricação de amortecedores.

Antes da aquisição pela Magneti Marelli, a Cofap já possuía um

Centro Técnico de Aplicação em Detroit e uma fábrica recém-

inaugurada na cidade de Kingsport, Tenessee, porém nesta época as demandas da

globalização exigiriam mais.


205

A meta é estar onde as montadoras se fazem presentes, e para isto, um plano de expansão

foi iniciado, com a implantação de fábricas na Polônia, Índia e China (estas duas via Joint-

Venture), além de mais um Centro Técnico de Aplicação em Torino, próximo da matriz.

3 Por que o PLM?

Em 1997 a MM Cofap já tinha um sistema de

Gerenciamento Eletrônico de Documentos (GED)

em início de implantação.

Em 2002 já não havia no então “arquivo técnico”

mais mapotecas, e os desenhos já eram

acessados via portal web.

3.1 Por que a Demanda então do PLM?

Porém é interessante também abordarmos um cenário que não é estranho quando se discute

implantação de sistemas informatizados para gestão de documentos, de projetos e

processos.

Muitas vezes há um volume razoável de investimentos, em hardware, software, consultorias

para atender as necessidades decorrentes do grande volume de informações que o ambiente

de negócios requer.

Estes recursos são disponibilizados para por fim a especificações duplicadas, diminuir a

quantidade de retrabalhos entre outros, porém é difícil a mensuração dos resultados após um

longo período de implantação de sistemas de gestão, e a culpa acaba sendo do software.

A MM Cofap também não escapou deste paradigma, e mais adiante, falaremos sobre o que

seria uma boa prática quando desejamos implantar software de gestão de Ciclo de Vida de

Produto.


206

A Cofap se tornou MM Cofap, deixou de ser nacional e agora como multinacional, com novos

Centros Técnicos, novas Plantas novas línguas é necessário transferir o conhecimento, uma

vez que a linha de negócios de amortecedores é a única do grupo com seu lead center de

desenvolvimento baseado no Brasil.

Neste momento os projetos tornam-se globais, e padronizar os processos é um requisito para

redução do tempo de desenvolvimento.

O Projeto PLM chega não só na MM Cofap,

mas em todo o Grupo FIAT, holding da qual a

Magneti Marelli faz parte para suportar estas

atividades.

4 Sobre a Implantação

Sair de GED para PLM é um desafio.

O implantador da solução e o antigo parceiro do sistema legado precisaram desenvolver uma

ferramenta para migrar os dados de um banco de dados para outro.

Foi elaborado um faseamento do Projeto, que apesar de ter sido apresentado pela equipe

italiana no final de 2004, somente foi iniciado o processo de importação no início de 2006,

quando o processo de migração foi definido.

O ano de 2006 não foi um ano tranquilo, enquanto um sistema era implantado, tínhamos um

sistema legado cujo servidor e banco de dados já não suportavam o volume de informações

e com frequentes paradas por falta de espaço em disco.

Seguem abaixo as principais fases definidas para o trabalho de 2006 até 2014:

2006 – Fase 1

Definição de parâmetros e criação programa de importação

Importação de dados e monitoramento

Implantação Consulta Worldwide

1ª versão workflow de documentos

2009 – Fase 2

Implantação Classification

2ª versão workflow de engenharia

Implantação workflow de solicitações de projetos


207

Migração Catia V5

Implantação Integração Catia V5 - Teamcenter

2013 – Fase 3

Implantação novo sistema de codificação

Implantação Structure Manager

2014 – Fase 4

Interface Teamcenter - ERP

4.1 A importância da Fase 2

Após a implantação da primeira fase, tínhamos finalmente encerrado uma plataforma de

software e iniciado outra, porém, após um ano de utilização, o estudo mais detalhado do

escopo mostrava que apesar de estarmos com um software para PLM, na realidade

estávamos com o sistema de Gerenciamento Eletrônico de Documentos mais caro do mundo,

pois a funcionalidade mais utilizada continuava sendo o controle de revisões e a consulta

eletrônica de documentação.

Foi contratada uma consultoria de PLM indicado pelo fornecedor de software, e revisamos as

funcionalidades mais necessárias, através de entrevistas com todas as áreas de P&D e

também com as áreas que faziam parte do processo de desenvolvimento do produto.

5 Estrutura Organizacional: Know-How Management

Para suportar e administrar o Projeto PLM, foi necessário revisar a estrutura organizacional

dentro da empresa.


208

Esta questão também faz parte de discussões em muitas organizações que começaram a

utilizar sistemas de Engenharia, como CAD, CAM, FEA, etc.

Quem administra? Quem implanta? É a área de TI? É um engenheiro, ou técnico com mais

aptidão para Tecnologia de Informação.

A criação do departamento de P&D, subordinado à Diretoria foi definido, utilizando-se como

base a mesma solução adotada pela Unidade de Suspensão da Marelli.

Uma equipe com analistas de sistema, alguns com experiência prévia em Engenharia, e

responsável por administrar os sistemas de gestão de engenharia foi formada, enquanto que

TI, fica responsável por suportar e validar as soluções de infraestrutura de rede, hardware e

adequação as regras de segurança da informação.

Obviamente, grande parte das atividades de Know-How Management é voltada ao suporte e

administração do sistema PLM.

5.1 Entregáveis do Projeto

Atrelado às entregas do Projeto, uma mudança de cultura de trabalho foi, e vem sendo

demandada.

Ao listar a seguir algumas das funcionalidades do software que são utilizadas, não pode ser

omitido que algumas mudanças não vieram para de imediato, economizar tempo, ou

simplificar a vida do usuário final.

Pode-se comparar isto à migração de um software CAD 2D para um CAD 3D, com o conceito

de assemblies, ou então as implantações de ERP, como SAP, que de início trazem um grande

impacto nos procedimentos de trabalho, mas que no longo prazo trazem o retorno.

Devemos interpretar os sistemas de gestão também como práticas de gestão do

conhecimento, enquanto que ao utilizarmos workflows, formulários eletrônicos, recurso de

pesquisa por atributos está estruturando um grande volume de conhecimento que estava fora

de um processo padronizado de trabalho.

Conscientizar os usuários do sistema sobre este período de adaptação é fundamental para

garantir aderência ao Projeto, no nível local e global.

5.2 Integrar CAD ao PLM

Sem dúvida, ainda um dos grandes desafios, ainda não temos 100% do processo de

modelamento integrado, ou seja, utilizando a interface do software PLM como a porta de

entrada para o trabalho no CAD, porém o upload de todos os modelos ocorre com o integrador

da plataforma, o que garante que o reuso destes modelos por outros centros técnicos

ocorrerá sem conflitos por perda de links ou erros de modelamento.


209

5.3 Classificação por Atributos

Utilizar a classificação por atributos foi um dos recursos do sistema priorizado, devido ao

grande potencial de reduzir efetivamente o lead-time de desenvolvimento do produto e dos

processos de cotação.

O amortecedor tem diversos componentes

que são utilizados em mais de um produto,

principalmente os conjuntos internos de

válvulas, e o registro das similaridades

geométricas e características de matéria-

prima, tem como grande foco o incremento do

reuso de peças e evitar a criação de novos

part numbers para componentes que

eventualmente já existam na base de dados.

5.4 Workflows e Auditoria de Processo

Automatizar um processo do negócio não é atualmente, algo inédito em grandes corporações.

Os processos de gestão de mudança, de acionamento de engenharia e diversos outros

pequenos acionamentos, cuja evidência dependia do e-mail, ou de um registro assinado,

foram substituídos por um workflow dentro do Teamcenter Engineering.

Porém, temos insistido, especialmente com os gestores de P&D, que um grande benefício

não percebido ao automatizar um processo, é o potencial de termos uma fotografia da “saúde”

e do desempenho do negócio.

Quanto tempo demora um fluxo de aprovação de desenho? Quantas rejeições têm em

determinada etapa?

Considerar o workflow sob esta ótica torna-se uma grande justificativa para sua implantação.


210

6 Engineering Bill of Material

A lista de peças de produto ou Engineering Bill of Material, hoje é criada no software de PLM.

Para fazer isto foi necessário um esforço de padronização, negociações com os

implantadores de SAP, que é o sistema ERP padrão do Grupo FIAT, e ainda demanda

melhorias, no que tange a uma melhor entrega para preparação da M-BOM (Manufacturing

Bill of Material).

A grande conquista na utilização da função Structure Manager foi a implantação de um

template global de produto para entregar as plantas produtivas.

6.1 Distribuição da Informação

Não existe mais em P&D um arquivo técnico, os fluxos de informação notificam via e-mail os

antigos receptores de informação em papel, que acessam o sistema, seja via web browser

ou via Portal para consultar o desenho, norma, procedimento ou relatório de ensaio, e se

necessário, imprimem para utilização.


211

6.2 Gestão Documental

A característica principal do PLM na MM Cofap é que o uso do sistema não está restrito

apenas aos desenhos da Engenharia de Produto.

Principalmente no Brasil, pela herança do trabalho em células multifuncionais, facilitaram um

intercâmbio com departamentos de qualidade, manufatura e produção.

Isto se reflete quando são gerenciados no banco de dados, além dos documentos de P&D,

também os desenhos de ferramentais, dispositivos da área de Manufatura, os relatórios dos

processos de APQP da área de Qualidade, além das normas de produção.

6.3 Conceitos que se Complementam e as Lições que Aprendemos

Em 2009 P&D conduziu um importante projeto de Gestão do Conhecimento, chamado de

Projeto SABER.

Este projeto mapeou os conhecimentos críticos e práticas de Gestão do Conhecimento

existentes e necessárias para toda a Diretoria de Pesquisa e Desenvolvimento de Produto.

Paralelamente ao iniciarmos o trabalho com a consultoria de PLM para implantarmos mais

funcionalidades existentes no sistema, identificamos que estas funcionalidades eram

respostas a muitas necessidades de compartilhamento, registro e proteção da base de

conhecimento existente em Amortecedores.

Foi a partir disto que Know-How Management utiliza toda oportunidade disponível, para, ao

treinar, identificar necessidades em PLM, trazer os conceitos de Gestão do Conhecimento

agregados ao sistema.

Fomentar uma cultura de Gestão do Conhecimento facilita o processo de implantação do

PLM, uma vez que dão argumentos junto ao público alvo, para entenderem que muitos frutos

serão obtidos também no legado futuro deixado aqueles que irão suceder os seniores que

estão na empresa, e facilitar no processo de enculturação para uma empresa global.


212

As lições aprendidas não terminaram, Projetos de PLM são projetos de alto risco, sob a ótica

de PMI: são projetos de longa duração, de alto valor de investimento e de escopo muito

amplo.

Precisam ser segmentados.

Muitas vezes, entregas pequenas, seguidas da mensuração do sucesso, municiam os

implementadores para justificarem mais recursos, considerando-se que ainda muitas

organizações consideram as áreas de P&D como custo indireto.

Um ex-diretor de P&D, italiano, cunhou uma frase a respeito disso: “Melhor um ovo hoje, do

que uma galinha nunca”.

Não é uma frase acadêmica, mas traduz muito do que devemos levar em consideração para

obtermos sucesso na utilização de sistemas informatizados no desenvolvimento de produto.

O apoio de um patrocinador forte, como na grande maioria dos projetos, é essencial. PLM

muitas vezes traz implicitamente a mudança de cultura.

Daí uma importante lição aprendida, é que devemos sempre levar em conta qual o grau de

maturidade da minha organização ao propor a utilização de determinada prática ou

funcionalidade de um sistema.

Uma funcionalidade cujo sucesso depende como pré-requisito de uma base completa de CAD

em 3D, se a engenharia ainda utiliza prancheta, suporta a decisão de postergar esta ação e

iniciar outra que foque a implantação de uma plataforma CAD.

Ao planejar os investimentos, devemos trazer junto à área de TI e levar em consideração os

custos agregados para uso da ferramenta.

Sistemas PLM, mesmo baseados na web, não rodam em computadores configurados para

acesso a aplicativos Office.

Se pensarmos em acesso global, banda de rede, servidores de replicação de dados, são

custos que não devem ser desconsiderados, para não extrapolarmos o budget de

investimento, e de despesa.

Finalizando, apesar de parecer óbvio, nem sempre isto acontece: o seu implantador, ou sua

área, ou mesmo a empresa, utilizam práticas e metodologias de gestão de Projeto, como do

PMI?

Esta é uma boa prática, o cronograma, o escopo bem definido, a gestão de tempo eficaz e a

documentação do projeto devem ser levados em consideração como fatores críticos de

sucesso para a implantação do PLM.


213


214


215


216


217


218


219


220


221


222


223


224


225


226


227


228


229


230


231


232


233


234

Documents

19 Seminario Internacional de Alta Tecnologia