Embed Size (px)
Citation preview
16-11-2012
1
Sistemas Multi-agentes na Indústria
Paulo Leitã[email protected]
http://www.ipb.pt/~pleitao
Seminário de Sistemas Inteligentes, Interacção e Multimedia
Porto, 15 de Novembro de 2012
Paulo Leitão, 2
Agenda
Contextualização de sistemas de produção
Sistemas multi-agente como uma solução para os requisitos da indústria
Discussão de aplicações de MAS na indústria
Análise das barreiras para uma maior adopção de MAS pela indústria e desafios futuros
Análise de um desafio emergente: integração de técnicas inspiradas na biologia
16-11-2012
2
Paulo Leitão, 3
Agenda
Contextualização de sistemas de produção
Paulo Leitão, 4
Sistemas de produção
● Tipos de processos produtivos:
- Manufactura (ou produção discreta)
- Processo contínuo
Processo de transformação que converte matéria-prima ou produtos semi-acabados em produtos finais e que possuem valor no mercado, usando operários e
maquinaria, e usualmente executada sistematicamente.
Fonte: M. Groover, “Automation, Production Systems and CIM”, Prentice-Hall, 1987.
16-11-2012
3
Paulo Leitão, 5
● Produção job shop
- Pequenas quantidades (frequentemente 1 unidade) de uma grande variedade de produtos.
● Produção por lotes
- Lotes de tamanho médio do mesmo produto.
● Produção em massa
- Especialização de um (eventualmente alguns) produto que possui uma procura elevada.
Tipos de produção: volume
Paulo Leitão, 6
Tipos de produção: layout
16-11-2012
4
Paulo Leitão, 7
Sistemas de automação
Paulo Leitão, 8
Sistemas de Fabrico Flexível (FMS)
sistemas convencionais detransferência
máquinas universaisconvencionais
Número de diferentes partes a serem processadas
Pro
du
ção
an
ua
l flexibilidade
produtividade
sistemas de fabrico flexíveis
Estrutura de produção que consiste num conjunto de estações de trabalho interligadas por um sistema de
transporte e manipulação de materiais, e controlado por um sistema computacional integrado.
16-11-2012
5
Paulo Leitão, 9
• Um sistema de produção consiste numa estrutura organizada de recursos físicos necessários à execução das funções de fabrico.
• Um sistema de produção será de pouca utilidade sem a presença de um sistema de controlo apropriado.
Supervisão e Controlo da Produção
Responsável pela execução física dos planos de produção, organizando, sincronizando e monitorizando o progresso
do produto que está a ser processado, montado, transportado ou inspeccionado na fábrica.
Paulo Leitão, 10
• O algoritmo de controlo deve decidir:
– Em que altura se deve produzir os produtos encomendados
– Que quantidade de cada produto se deve produzir
– Como e quando usar os recursos para produzir os produtos
• Os sistemas de fabrico são caracterizados por serem:
– Sujeitos a pressões dos mercados que procuram produtos customizados com reduzido prazo de entrega
– Sujeitos a perturbações, e.g. atrasos e avarias
– Não-lineares, complexos e caóticos
Complexidade do sistema de controlo
16-11-2012
6
Paulo Leitão, 11
Componentes do sistema de controlo
planeamento
despacho
monitorização diagnóstico
recuperação
plano de produção
escalonamento detalhado
informação para o planeamento
sistema de fabrico flexível
detecção de erro
comandos para os actuadores
sinais dos sensores
recomendação de estratégias
medidas de desempenho
dados tempo real
comandos para os actuadores
escalonamento
Paulo Leitão, 12
• São desenvolvidas e adaptadas caso a caso.
• Possuem fraca rapidez de resposta a perturbações.
• Necessitam de um grande esforço, para expandir, reconfigurar ou manter a aplicação de controlo.
• Não são adequadas porque não suportam eficientemente os requisitos actuais impostos aos sistemas de fabrico.
As abordagens tradicionais
16-11-2012
7
Paulo Leitão, 13
The industry requirements
video reconfiguration 1
mass customization with shorter product life
cycles
flexible and reconfigurable
production plants
more complex systems,
exhibiting re-configurability,robustness and responsiveness
what the market demands
what the companies needs to have
what are the challenges for the
system’s developer
video reconfiguration 2
Paulo Leitão, 14
Agenda
Sistemas multi-agente como uma solução para os requisitos da indústria
16-11-2012
8
Paulo Leitão, 15
• Conjunto de entidades distribuídas.
• Cada entidade é independente, possuindo:
− Objectivos, conhecimento e competências próprias.
− Um conjunto de regras que regula o seu comportamento.
• Nenhuma entidade tem acesso a toda a informação.
• As decisões são determinadas através da interacção entre mais do que uma entidade.
• As entidades podem estar ligadas a dispositivos físicos.
As novas abordagens devem …
Paulo Leitão, 16
• Elevado nível de autonomia e cooperação, não existindo a estrutura cliente-servidor com relações fixas.
• Agente: Componente autónomo que representa objectos físicos ou lógicos do sistema, capaz de actuar de forma a atingir os seus objectivos
• Outras características: reactividade, pró-actividade, habilitações sociais.
• Sistema multi-agente: Conjunto de agentes capazes de interactuarem de forma a atingirem os seus objectivos individuais, quando não possuam conhecimento
e/ou competências para os atingirem de forma independente.
Sistemas Multi-agente (MAS)
16-11-2012
9
Paulo Leitão, 17
MAS working in practice
visão local
Comportamento local
Paulo Leitão, 18
What multi-agents can offer
Reusabilityold components can be re-used to develop new components or new systems
Distributed thinkinga complex problem can be divided into several small problems
Modularitybuilding the system by
pieces like using LEGO
Robustnesslosing one decision node doesn’t implies the system failure
Reconfigurabilitychanges can be performed on the fly
Smooth migrationfrom old technologies to new ones
16-11-2012
10
Paulo Leitão, 19
Exemplo de aplicação
PLCIEC 61131
Holon Transporte
Holon Máquina #1
CNCWin NT appl.
Holon Máquina # 3
Holon Máquina #2
Não, estou sobrecarregado!
IPCWin NT appl.
Holon PeçaQuem pode furar esta peça?
Estou fora de serviço!
Sim, eu posso!
Eu transporto a peça!
Quem a pode pode transportar?
Paulo Leitão, 20
Comparação entre CIM e HMSAbordagem Convencional (CIM) Abordagem distribuída e inteligente
(HMS)
Abordagem top-down Abordagem bottom-up
Solução centralizada para cada função individual de controlo
Solução distribuída com cooperação entre nós e simultaneamente focando mais do que uma função
de controlo
Arquitectura rígida e estática Arquitectura flexível, programável e dinâmica
Relações cliente-servidor Relações holon-holon
Comunicações um para muitos (1-N) Comunicações muitos para muitos (N-M)
Inteligência concentrada nos níveis de topo Inteligência distribuída pelos níveis de controlo
Eficiência através da especialização Eficiência através da flexibilidade
Fraca resposta a perturbações Elevada resposta a perturbações
Mais eficiente para elevado volume e pequena variabilidade
Mais indicado para elevado-baixo volume e média e alta variabilidade
Os operadores são substituídos por tecnologias de automação (retirados do processo produtivo)
Os operadores são complementados com tecnologias de automação (aumentando as competências dos operadores que ficam no
processo produtivo)
16-11-2012
11
Paulo Leitão, 21
Agenda
Discussão de aplicações de MAS na indústria
Paulo Leitão, 22
• Application in a wide range of domains:
– Electronic commerce, graphics (e.g., computer games and movies), transportation, logistics, robotics, manufacturing, telecommunications, energy, etc.
• Types of applications:
– Industrial applications, R&D projects, laboratorial prototypes
• Inside of the manufacturing domain, different levels:
– Supply chain, planning, scheduling & control, machine controllers.
Initial considerations
16-11-2012
12
Paulo Leitão, 23
• Software developers companies
• Automation technology providers
Companies offering MAS solutions
Paulo Leitão, 24
Factory producing engines: Daimler Chrysler
• Application in a factory plant of Daimler Chrysler
• Production of cylinder heads for four-cylinder diesel engines (used in the Mercedes Benz C and E class 220 CDI)
• Objective: meet rapidly changing operations targets
16-11-2012
13
Paulo Leitão, 25
Initial structure of the system
Buffer ProductionUnit
BufferProductionUnit
Paulo Leitão, 26
Consequences of the fixed structure
• Linear production
– Whole line will stop if one machine fails
– Availability of machines about 98 % but of line about 50...70 % only
– Dedicated buffer in between lines needed
• Long time for design and installation
– Requests to machine builders starts 3 years before lot #1
– Planned production quantity never hits (reality 25% - 300%), leads to waste of investment or lost sales
16-11-2012
14
Paulo Leitão, 27
Inflexibility of the system
• Scheduling fixed and “built in” part design– No use of lines for other parts
Camshaft Crankshaft
Paulo Leitão, 28
Need for a new structure
• Specific Machines
– No flexibility for short, medium and long term changes of lot sizes
– No reuse of machine components for modifications
• Paradoxon of Load
– Cylinder head production needs 490 spindles, drilling/milling 3,95kg steel,
o 45 spindles cover 80% (average 67 g / spindle)
o 445 spindles cover 20% (average 2 g / spindle)
16-11-2012
15
Paulo Leitão, 29
Agentification of the hardware system
Machine Machine
Loader
Unloader
Shift table
workpiece
Ethernet TCP/IP
agente peça
agente máquina
agente transporte
SW/HWDesign
Paulo Leitão, 30
Some technical details
• Agent-based system (Production 2000+) to allow individual workpieces to be directed dynamically around the production area.
• Agents to represent:
− machines and work-pieces
• Dynamic resource allocation using a CNP -based schema (objective: optimization of the throughput).
• Redundancy gives the possibility of diverting product to another machine if a breakdown or unavailability occurs.
16-11-2012
16
Paulo Leitão, 31
Some results
4:00PM
5:00PM
6:00PM
7:00PM
8:00PM
time
Machine 1
Machine 2
Machine 3
0
2
4
6
8
10
12
14
16
18
20
6:00AM
7:00AM
8:00AM
9:00AM
10:00AM
11:00AM
12:00PM
1:00PM
2:00PM
3:00PM
Machine 3 takes over jobs of machine 2 Machine 2: Tool broken Machine 1 is “plugged into”the system with auto configuration
• Day-to-day operation for five years up to the end of the life-cycle of the targeted product.
• 20% increase in productivity on average.
Paulo Leitão, 32
• Holonic control of an assembly system in the automotive industry, namely for assembling engines.
• Technical details:– Holons for each docking station, engine buffer, machine station
and AGV.
– Uses, as the P2000+, a CNP to request resources, e.g., AGVs.
• Differences to P2000+ control system:– System design: the production of cylinder heads is different from
the assembly of engines.
– Expected requirements: flexibility and robustness for the P2000+, robustness and scalability for the Holomobiles.
Holomobiles
16-11-2012
17
Paulo Leitão, 33
HVAC systems: US Navy ships
• Application: − Agent-based control system for
the chilled water systems and the heating, ventilation and air conditioned (HVAC) systems of the US Navy ships.
• Objective:
• Plan, commit and execute control tasks.
Paulo Leitão, 34
Some technical details
• Follows the FIPA specifications.
• Agents represent:
− Physical devices (T-pipe, cooling plants, water services and heat loads).
− Specific ship functions (chilled-water, material-handling, heat and ventilation, combat subsystems, …).
• The agents reside on the PLCs.
• CNP to establish dynamic negotiations among the agents.
16-11-2012
18
Paulo Leitão, 35
• Holonic control system to assembly Gillete™ packages into customer-tailored gift boxes.
• 2 Fanuc M6i robots, a storage system and a conveyor system.
• Technical details:
– Considers order holons and resource holons that represent the physical components of the system.
– Uses the JACK Intelligent Agents™ platform.
– Integrates RFID technology, by using electronic tags embedded in discrete units, replacing the traditional barcodes.
Cambridge packing cell: Gillete
video cambridge 1
Paulo Leitão, 36
• Objective: develop an agent‐based simulation tool aimed at tasks of dynamic product routing.
• Technical details:– Agents, representing transportation components, negotiate about
optimal routes in a redundant conveyor system.
MAST (Manufacturing Agent Simulation Tool)
– Find alternative routes if failures detected or changes in the layout.
– Implemented using JADE.
– Interface to PLCs, enabling the control of a real system.
16-11-2012
19
Paulo Leitão, 37
• MAS using an ant-based strategy (developed by NuTech Solutions) to reduce production and distribution costs, namely by:
– Optimizing the truck routes for delivering industrial and medical gases
– Adapting the production schedules to changing conditions
Air Liquide America
Paulo Leitão, 38
• Southwest Airlines is using ant-based algorithms at Phoenix Sky Harbor Airport to get planes to available gates faster.
• The program was so successful that Southwest is now applying ant algorithms to the ticketing and check-in process.
Sky Harbor International Airport in Phoenix
16-11-2012
20
Paulo Leitão, 39
• A multi-agent system combined with swarm intelligence principles were used to forecast energy demands in Turkey until 2025.
Forecast of energy consumption
• Each agent acts like an ant while foraging for food.
• This ant colony optimization based algorithm has proved to behave better than others when comparing results from previous years.
Paulo Leitão, 40
• Tankers International operates one of the largest oil tanker pools in the world.
• Objective: optimize the management of large-capacity tankers, carrying out transcontinental transportations of oil using a very large crude carrier fleet.
• An agent-based optimizer, Ocean i-Scheduler, dynamically schedules in real-time the cargo assignment to vessels in the fleet.
Intelligent Fleet Cargo Scheduling: Tankers International
16-11-2012
21
Paulo Leitão, 41
• Agent-based optimization techniques provides:– Increase of the responsiveness, by adapting plans in response to
unexpected changes, such as transportation cost fluctuations or changes to vessels, ports or cargo.
– Reduction of the costly mistakes, and preserving the knowledge developed in the process of scheduling.
• A new order can affect changes of a lot of tankers and even alteration of contracts with a number of clients.– "By modeling each tanker as an individual agent is achieved the
ability to see the options and ability to respond rapidly to emerging events in real time.“ Michael Luck
• Cost of one day of idle time of each tanker is $100,000.
More details
Paulo Leitão, 42
• MAS dynamic real time scheduling system for rental cars– Across 100 stations, 15.000 cars in UK
– Cutting costs of secondary logistics for car-repositioning (e.g. wages for drivers, overnight payments and travel expenses)
• Benefits:– Decrease of idle miles by 23 %
– Increase of driver utilization by 17%
– Increase of the fleet utilization
– Savings on fuel expenses and drivers in overtimes
AVIS rent-a-car
16-11-2012
22
Paulo Leitão, 43
• Real-time planning and optimization of a fleet of taxis operations
Addison Lee taxis corporation
• Benefits:– Automated scheduling 97% of 13.000 orders for 800 taxis per day
(dispatchers became supervisors)
– Response-failures reduced by factor of 3.5 (2%)
– Processing-efficiency of orders improved by 60%
– Average profitability per taxi increased by 5 %
– Utilisation improved by > 20 %
– Training costs for new operators reduced by 75%
Paulo Leitão, 44
• Real time collection of hundreds of streams of data and more than a billion of data points in a race weekend
• Looking to get beyond predictive intelligence to prescriptive intelligence
• When it is not possible to predict it is crucial to adapt on the fly.
McLaren applied technologies
16-11-2012
23
Paulo Leitão, 45
• Autonomous learning agents for decentralised data and information networks
• Mechanisms, architectures, and techniques to deal with the dynamic and uncertain nature of distributed and decentralised intelligent systems.
• Application: decentralised coordination in RoboCup rescue
ALADDIN project
Paulo Leitão, 46
• Use autonomous agents to balance multiple demands in a satellite,
– Such as staying on course, keeping experiments running, and dealing with the unexpected, thereby avoiding waste.
• Besides responding to predefined events, agents
– Can react to unimagined events .
– Still ensure that the spacecraft does not waste fuel while keeping to its mission.
NASA satellites
16-11-2012
24
Paulo Leitão, 47
• Agent technology was used to model individual combatants in the Peter Jackson’s trilogy The Lord of the Rings.
Movies: The Lord of the Rings
Paulo Leitão, 48
Relevant Ancient EU R&D projects
Application Description
MASCADA (Valckenaerset al., 1999)
Manufacturing control mechanisms to support the production change and disturbance, safeguarding and/or maximizing the production systems throughput; uses autonomous and intelligent agents to represent the factory components.
IntaPS(Denkena et al., 2002)
The presents an approach for integrated process planning and production control, which architecture consists of two main components, which link information systems of earlier stages of product development and the resources on the shop floor. This link is realized by decentralized planning on shop floor level and by rough level process planning.
PABADIS (Sauter and Massotte, 2001)
Uses the concept of CMUs (Co-operative Manufacturing Units) to provide functions to the production process in automation control, encapsulating residential, products and shop floor management as agents; comprises centralized (for the connection with ERP systems) and decentralized components, being the products implemented using the mobile agent technology.
16-11-2012
25
Paulo Leitão, 49
On-going EU R&D Projects
Title Short description
GRACE Use of multi-agent systems for integrating process and quality control, and consider self-adaptive procedures into control and diagnostic systems at local and global level. Demonstration considers a washing machine production line.
IDEAS Development of demonstrators/technological solutions that proves that assembly equipment can be highly adaptable, applying the concepts of EAS.Focus in the agent-based fault-tolerant control and reconfiguration aspects.
COSMOS Design, development and implementation of control systems with a flexible, modular and evolvable automation approach, increasing productivity without losing flexibility.
COLLIS.EUS
Development of collaborative information systems involving multiple interacting agents and soft-computing techniques for robotic and sensor systems.
CONET Development of a community capable of conducting the research to achieve the vision of combining embedded systems for robotics and control, pervasive computing and wireless sensor networks.
Self-learning Uses highly reliable and secure service-based self-learning solutions aiming the integration of control and maintenance of production systems.
Paulo Leitão, 50
• Objective: Develop and validate a collaborative MAS which operates at all stages of a production line.
• Aligned with the need to build modular, intelligent and distributed manufacturing control systems.
• Focused in product-driven production.
• Adaptation/optimization of production and quality control processes, at the level of local agents.
• Application at a real industrial washing machines plant.
16-11-2012
26
Paulo Leitão, 51
PA PA PAPA PAPA
QCA QCA RA
. . .
IMA
provides the new parameters to be written
provides results from quality
control
provides results from quality
control
Adaptation in GRACE MAS production system
Paulo Leitão, 52
• Title: Adaptive Production Management
• Focus: adaptive ramp-up management in the production of highly customized products, especially small lot sizes.
ARUM
16-11-2012
27
Paulo Leitão, 53
• Objective:
– Develop mitigation strategies to respond faster to unexpected events in production and ramp-up of complex and highly customized products.
– Based on a new generation of service orientated enterprise information platforms, a service oriented bus integrating SoA and knowledge-based MAS.
• The solution integrates multiple layers of sensors, legacy systems and agent-based tools for beneficial services like learning, quality, risk and cost management.
• Start date: September 1st, 2012 (duration: 37 months).
ARUM
Paulo Leitão, 54
Consortium
16-11-2012
28
Paulo Leitão, 55
Current status
• Major focus of development and deployment is centered in Europe
• Very small adoption of agents in industry
• The implemented applications are limited (in terms of functionality)
• Addresses the high-level control or pure software systems (e.g. electronic commerce)
• Few enthusiasm from both the technology providers and the industry companies
Paulo Leitão, 56
Agenda
Análise das barreiras para uma maior adopção de MAS pela indústria e desafios futuros
16-11-2012
29
Paulo Leitão, 57
IEEE TC on Industrial Agents
http://www.tcia.ieee-ies.org/
Paulo Leitão, 58
Mission of the TC on Industrial Agents
• Provide a forum to exchange ideas, knowledge, experience, learning and results in this area of expertise.
• Stimulate contacts and establish links with and between industry and academia to drive industrial agents.
contributing for a wider application of industrial agent technology in distributed production, services and infrastructure
sectors
16-11-2012
30
Paulo Leitão, 59
Road-blockers
investment
distributed thinking interoperability
scalability standardization real-time constraints
integration with physical devices supporting
technologies and methodologies
Paulo Leitão, 60
Current trends
• Convince industry people of the benefits of using agents.
• Examples of actions:
– Providing demonstrators running in industry
o Showing the maturity, flexibility and robustness of the technology.
o Allowing companies to “believe” in the agent technology.
– Provide ROI analysis
o Considering the development + operation + maintenance costs.
– Provide agent-based solutions as black boxes
o Hiding the system complexity by providing interfaces and configuration tools.
o Analogy to our cars and washing machines.
16-11-2012
31
Paulo Leitão, 61
Standardization
• Standardization issue is pointed out by industry as a major challenge for the industrial acceptance of the agent technology
• Standards may affect the development of industrial MAS solutions, namely:
– FIPA, IEC 61131-3, ISA 95, Web services, semantics (ontologies, OWL, …)
• Standardization should be seen in two different perspectives:
– Fulfil the current related industrial standards
– Introduction of new standards or influence the existing ones
Paulo Leitão, 62
Integration of other concepts and technologies
• Integration with other complementary technologies, e.g.
– IEC61131-3 and IEC 61499 standards to implement the low-level control that is not addressed by the agents
– Service Oriented Architectures (SOA) / Web services to solve the interoperability problems allowing the vertical and horizontal integration.
• Integration of bio-inspired techniques and methods.
16-11-2012
32
Paulo Leitão, 63
Agenda
Análise de um desafio emergente: integração de técnicas inspiradas na biologia
Paulo Leitão, 64
Engineering of distributed systems
InfrastructureTechnologies
(Wireless sensornetworks, RFiD, ...)
Collaborativecontrol paradigms
(HMS, MAS, SoA, ..)
Biological inspired Techniques
(Self-organization, Emergent Behavior, Swarm
intelligence, ..)
Questions:• How the global
optimization is achieved?
• How holarchies are dynamically formed, evolved and removed?
• How individual components self-organize and evolve?
• How to adapt their emergent behavior using learning?
16-11-2012
33
Paulo Leitão, 65
What nature and biology offers
Complex Adaptive
Engineering Systems
Plenty of powerful mechanisms to handle complex environments
Complex systems built upon entities exhibiting simple behaviors
Examples: Air Liquide, Sky Harbor Airport, …
bird flocking
ants foraging
fish schooling
Paulo Leitão, 66
• Complex systems arise out from a multiplicity of interactions among entities exhibiting simple behavior
• Micro-macro effect:
– Micro level, considering the system from the point of view of the local components and their interactions
– Macro level, considering the system as a whole being the result from the lower-level interactions
What is the meaning of Emergence?
16-11-2012
34
Paulo Leitão, 67
• Individual organisms use relatively simple behavioral rules
• Achieved behavior and patterns are more complex than the individuals from which they emerge
• Complexity comes additionally from:
– Sensitivity to initial conditions (butterfly effect)
– Non-linear interactions among components involving amplification and cooperation
Complexity in Emergence
Paulo Leitão, 68
Emergence in Complex Systems
simple behaviors composed by a set of few simple rules
local behavior
p3
p4
p5
p1
t1
t2 t3
t4
p2
local behavior
p3
p2
t2
t3
p1
t1
p6
p5
t5
t6
p4
t4
p9
p7
p8
p3
p2
t2
t3
p1
t1
p6
p5
t5
t6
p4
t4
p9
p7
p8
local behavior
p1
p2 p4
p3 p5
p6
t1
t2 t3
t4
t5
local behavior
local behavior
p3
p4
p5
p1
t1
t2 t3
t4
p2
p2 t2p1 t1
emergence:
the whole system emerges from the
interaction between local entities
positive feedback (amplifying)
negative feedback (dampening)
complex system behavior
16-11-2012
35
Paulo Leitão, 69
• Systems exhibiting the emergent behavior are:
– Flexible: adaptation to changing environments by adding, removing or modifying the entities on the fly
– Robust: society of entities has the ability to work even if some individuals may fail to perform their tasks
• But a whole behavior that is difficult to predict: – Large number of non-linear interactions large number of
possible non-deterministic ways the system can behave
• Desirable to ensure: – Expected properties will actually emerge
– Not expected and not desired properties will not emerge
System Exhibiting Emergent Behavior
Paulo Leitão, 70
Self-organization
• MAS usually misses truly self-adaptation
• Process of evolution where,
– The development of novel, complex structures takes place primarily through the system itself
– Normally triggered by internal variation processes, which are usually called "fluctuations" or "forces”
fish
scho
olin
gbi
rds
flock
ing
16-11-2012
36
Paulo Leitão, 71
Self-organization in several fields
birds flocking
Hanoi traficShibuya crossing
antschemical reaction
Paulo Leitão, 72
• A form of self-organization, involving an indirect coordination between entities
– The trace left in the environment stimulates the execution of a subsequent action
– By the same or different entity
Stigmergy
16-11-2012
37
Paulo Leitão, 73
• BiosGroup buys NuTech Solutions, Inc.
• BiosGroup• Founded in 1996 by S. Kauffman and the Center for Business Innovation of Ernst &
Young, is the world leader in applying the science of complexity and complex adaptive systems to the simulation, modeling, and solving of difficult problems.
• Declarations:
– “The combination of BiosGroup and NuTech Solutions will accelerate the commercialization of the science (of complexity) and will result in more products that use the science to solve problems facing decision makers“ S. Kauffman.
– “BiosGroup has been rich in the technology of agent-based modeling and simulation solutions that fits in well with NuTech Solutions' predictive analytic and profit optimization software. We are going to be part of a very interesting company“ R. MacDonald.
Bio-inspiration + MAS
Paulo Leitão, 74
Thank you!
e-mail: [email protected]
URL: http://www.ipb.pt/~pleitao
