images

Smart Decisions in Complex Systems

Pierre Massotte

Patrick Corsi

image

Preface

Why do we need to look at complexity?

When complexity is a part of everyone’s daily experience, what is more fitting than a book that aims to process this “complexity”? When catastrophes of every kind appear on the media screens in our homes, it may be useful to question the true meaning of the word “catastrophe”. According to some, the term means chaos and disorganization; it can also mean, for example, the return to rest of a vibrating musical string as per the laws of mechanical resonance. Likewise, it may be equally useful to contemplate on the true meaning of the word “chaos”: in the beginning chaos is not differentiated from divine thought and is, in a way, the matrix of a future yet to become, as well as an opening on new ways of thinking.

In order to better understand the scope of this issue, it is worth recalling the historical approach as employed by Science since the 17th Century, when Descartes published the “Discourse of Method”, which serves as the foundation of modern rationalism and its ongoing development. From a scientific point of view, this “classical” way of thinking is based on the fact that the world is a rational, mathematical, knowable and decomposable quantity. On the literary level, an examination of “classical” dramaturgy reveals the rule for three key devices (time, place and action). Notable playwrights whose oeuvres follow this stagecraft doctrine are Boileau, Corneille, Racine, etc. Essentially, these principles advocate that everything can be systematized, decomposed and organized, and is recognized as the basis for the great progresses in knowledge and the management of systems. They continue to permeate the overall scientific approach, which is characteristically fragmented, isolated and centralized.

However, this concept and essence of Truth reaches its own limitations, as it inevitably leads to a hierarchical organization that limits our abilities and potentials for research and engineering, at times juxtaposed, and is responsible for creating inherently fixed spaces with reduced freedoms, which deny multidisciplinary cooperation, consultation and synergy. As such there are many opportunities for development and problem resolution.

Hence, the cultural heritage of the past has not predisposed us to the current socio-economic transformations created by globalization, as it is applied to other contexts and approaches, and which is already changing towards a single-system approach, often referred to as the holistic approach.

What does this book claim?

The contribution of Complexity Science is, in a sense, an attempt to rebalance the classical analytical approach and its particular limitations. This can be seen to be made up of individual perceptions that together become the complementary polar, thereby allowing for a global understanding of the world, our systems and our societal behaviors.

While this complement has become as important as the Cartesian approach (i.e. the analytical approach) to our context, it has also become increasingly urgent. Everything around us pushes us to review our patterns, to enlarge and to stretch them to the point of breaking our educational, behavioral and structural limits. It has become essential to understand the Global and to propose a new path based on connectionism and self-organization. The Global is before us, within us, within our reach and on our scale; it is the awareness of factors that we have so far hidden from ourselves. “How long does a fly live?” A life. And Pieng-Tsou, the oldest man in the world? Also a life”. (Shipper). It is our deafness to the relationships between entities and the factors that connect phenomena, which has unintentionally limited science and engineering so far.

This book proposes to open a door. Without negating all the progress made so far, it is the authors’ belief that the time has come to give prominence to a conscious and reasoned apprehension of the Global and the role it plays in our socio-economic lives. Our aim is to introduce a new paradigm, which we have experimented on with concrete case studies, and to establish a preliminary set of scientific and technological bases. That said, this book does not aim to be a theoretical or scientific contribution and is rather intended for all those wishing to broaden their practice of management and engineering systems. As such it has been written for engineers and technicians, strategists and planners, managers, researchers, teachers, and students. It is but a first step, carrying with it the hope that it might inspire hitherto unknown advancements, as well as engage other authors to grapple with this valuable appreciation of complexity and its many applications on the ground.

By writing this book…

The history behind the genesis of a work is sometimes most unexpected. Since the 1970s, both of the authors have been situated within the industry, and have been involved with the creation of new technologies, privileged to experience each phase of the great wave that is artificial intelligence. One of the authors has spearheaded multiple projects investigating the limitations of knowledge, research that was conducted at IBM France, for Networks and Telecommunications at the La Gaude Research and Development Center, for large-scale computer systems at the Pompignane plant near Montpellier, and finally, with IBM Europe in charge of research and development projects to improve the competitiveness of the group’s R&D plants and centers. As such, over the latter part of the last century, he has coordinated large teams comprised of more than 60 people. The other author was formerly employed at IBM’s San Jose Research Laboratory in California, then at the La Gaude Research and Studies Center, after which he became involved with an artificial intelligence start-up, and ran the R&D management for a subsidiary of THOMSON-CSF, where he oversaw the portfolio for advanced artificial intelligence projects in line with the European Commission in Brussels. As such he is reputed as an expert member on numerous European commissions concerned with complex approaches, and is a specialist in designing breakthrough innovations pertinent to the complex processes involved.

Neither author came across the Complexity Sciences by chance. After having met over the application of neural software networks in the late 1980s, it was not until the 1990s when the “era of networks” became irrevocably established, and with it the associated frustration brought about by increasingly complicated computing systems, that the real connection happened. Their paths then separated, with one choosing to terminate a long career as Head of Research at IBM in the field of Complexity and the Transfer of Technologies in Industry at the Ecole des Mines d’Alès (EMA) in Nîmes; and the other embracing a career with the European Commission, and later working as an international contractor and consultant. At this pivotal point, they were struck by the acceleration of transformations within the industry, and became convinced that it was mostly due to the quality of interrelations between previously isolated elements. This progression, they felt, would need to mature in the years to come. As privileged observers, users and internal actors of the various instruments implemented in the planning and conduct of European research and development, and framework programs for more than 20 years, the authors noted a growing incongruity, the novelty without appeal of conventional systems strategies, the limitations of top-down planning and monitoring. A new dynamic in the markets appeared by way of a transdisciplinary sidelong perspective. Success stories emerged less from structured bureaucracies and more from the mass market, a growing arena that connects all the actors in a hectic agora with a seemingly irrepressible capacity for innovation. In short, systems were no longer the solution. They had become the problem.

In terms of networking the citizens of the world, any organization not able to operate at the level of intensity and adaptation of its actors is rendered null and void. But how to explain this general sense to our customers? How can they free themselves from the obstacles to their own innovation, from the structural obstacles to their competitiveness, often generated by themselves in an earlier era? Little by little we had the same idea, to write this book, to bring forward the elements of a strategy for change and to make these accessible to all organizations and companies.

With the good fortune of a rich research heritage of the first magnitude, spanning more than 30 years, we have developed a method for developing and launching “global applications”. The intention was not to develop an academic work, but rather to focus on setting up methodological bases, validating and refining this new discipline, as consultants and entrepreneurs, to play our part in the global management practices of companies, organizations and consultancies, to help managers identify and model, internalize and innovate, in a word, to experience the crucial decisions tied to their “global apps” of tomorrow. Without a comprehensive approach, there will only be failures and ruination on the path to business success. The maxim “Think globally to act locally” led to our own motto: “Formulate globally to decide locally”.

Having interviewed managers, companies and administrations, we examined their organizational structures and their successes and failures: whether in production, distribution and indeed throughout the supply chain; in financial institutions, administrative and social institutions, as well as research organizations. We also examined the dynamic links of these managers and how they interact with their associates and partners, suppliers and customers, their structures, their business model: in short, their economy. The job of tomorrow is there because of these relationships. It will be the wealth of the old nations too; bearing in mind we always observe through the appropriate lens. It is important to change our thinking, even if it involves a shift in our cultural, organizational and economic paradigms.

Pierre MASSOTTE
Patrick CORSI
April 2017

Acknowledgments

When writing these first books, we received many suggestions from friends and colleagues. The formation of ideas was inspired from firsthand experience in the field. These ideas were enriched by crossfertilization and discussions held with researchers from the IBM Corporation and the Research Institutes in France, notably at the Ecole des Mines d’Alès (Henri Pugnere, I.G des Mines, and Gérard Unternaehrer, I.G Arùmmement) and abroad (such as the Santa Fe Institute). Many thanks to our former colleagues at the European Commission, who, in addition to their already heavy burden, made themselves available and who were very generous with their attention. We equally benefited from the commitment of friends and colleagues to clarify lesser known points of their strategic planning, with several of them allowing us access to information in order to establish concrete cases.

This book, which is at the same time sourced from these two preceding publications, as well as constituting a significant increase in order to reflect the socio-technological developments of our time, could not exist without the help and support of co-workers and management. As such we would like to take this opportunity to express our thanks to: Jean Taverne, General Manager of Technical Services of IBM France who carried out experiments on IBM France sites and helped to set up the former European Competencies Center in Artificial Intelligence; René Balmès (IBM Global Services) who was a great visionary in the management of complex systems; Scott Kirkpatrick, Benoît Mandelbrot and John Sowa of the IBM Research Division in Yorktown Heights, NY, who always responded to requests for information. Finally, from the academic point of view, thanks must go to the teachers Pierre Ladet (Grenoble) and Alain Haurat (Annecy), as well as Prof. Hermann Kuhnle (F.I. Magdeburg) and Prof. Abdelhakim Artiba (MonsUniversity) – who consistently encouraged the underlying works and inspired our confidence.

Thank you to ISTE for their support, patience and dedicated enthusiasm for the manuscript.

Finally, thank you to Anne Marie Massotte for helping within the completion of this book.

List of Acronyms

ACCA:
Agent-Container-Communication-Auto/Self-Organization
AFI:
Agri-Food Industry
AI:
Artificial Intelligence
ANN:
Artificial Neural Networks
ATG:
Advanced Technology Group
B2B:
Business-to-Business
B2C:
Business-to-Consumer
BA:
Broker Agent
BDIN:
Belief, Desire, Intent, Need
BoM:
Bill of Materials
BPR:
Business Process Reengineering
BTO:
Build To Order
BTP:
Build To Program
CA:
Cellular Automata
CAD:
Computer-Aided Design
CAM:
Computer-Aided Manufacturing
CAP:
Computer-Aided Production
CAPM:
Computer-Assisted Production Management
CBR:
Case-Based Reasoning
CC:
Collaborative Consumption
CEA:
Commissariat à l’Energie Atomique
CFM:
Continuous Flow Manufacturing
CIM:
Computer-Integrated Manufacturing
CLT:
Central Limit Theorem
CML:
Complex Mutual Logistics
CMU:
Cooperative Manufacturing Unit
CNP:
Contract Net Protocol
COBOT:
Cooperative Robot
CSR:
Corporate Social Responsibility
DAIS:
Decision-Aid Interactive Systems
DAPS:
Dynamic Analyzer of a Production System
DE:
Differential Equations
DFT:
Demand Flow Technology
DLF:
Direct Line Feed
DP:
Dynamic Pricing
DSS:
Decision Support System
EMA:
Ecole des Mines d’Alès (France)
ERP:
Enterprise Resource Planning
FBL:
Feed-Back Loops
FBM:
Field Bills of Materials
FFT:
Fast Fourier Transform
GNOSIS:
Knowledge Systematization – Configuration Systems for Design and Manufacturing
IBM:
International Business Machine Corporation
IDAS:
Interactive Decision-Aid System
IDE:
Integral Differential Equations
IDSS:
Interactive Decision Support System
IIE:
Institute of Industrial Engineers
IMS:
Intelligent Manufacturing Systems
IOT:
Internet Of Things
IS:
Information System
JIT:
Just In Time
KADS:
Knowledge Acquisition and Data Structure (a project)
KBS:
Knowledge-Based Systems
LBD:
Ligand-Binding Domain
LCM:
Life-Cycle Management
MAQ:
Maximum Allowable Quantity
MAS:
Multi-Agents Systems
MCA:
Multiple Correspondence Analysis
MES:
Manufacturing Execution System
MFG:
Mean Field Games
MFG Order:
Manufacturing Order
MIMD:
Multiple Instruction on Multiple Data
MLP:
Multi-Layer Perceptron
MMI:
Man–Machine Interface
MPP:
Master Production Plan
MPS:
Master Production Scheduling
MRP:
Material Requirement Planning – Also: Material Resources Planning
NAN:
Nonlinear Adaptive Networks
NANN:
Nonlinear Adaptive Neural Network
NCP:
Neighborhood Coherence Principle
NICT:
New Information and Communication Technologies
NLAS:
Nonlinear Adaptive Networks
NLDS:
Nonlinear Dynamic Systems
NMPP:
New Manufacturing Production Paradigm
NP:
Negotiation Protocol
NPDI:
New Product Development and Introduction
ODE:
Ordinary Differential Equations
OR:
Operations Research
P2P:
Peer-to-Peer (or Point-to-Point)
PDE:
Partial Differential Equations
PLCs:
Programmable Logic Controllers
PLM:
Product Lifecycle Management
PLOOT:
Plant LayOut Optimization
PnP:
Plug-and-Participate
PPB:
Parts Per Billion
PPC:
Pull Production Control
PPM:
Parts Per Million
PR:
Production Reservation
QUETA:
European ESPRIT 4 project #22367 “Quality Engineering Tools for Assembly and Small Batches Manufacturingˮ
RFID:
Radio Frequency Identification Devices
RMLP:
Recurrent Multi-Layer Perceptron
ROI:
Return On Investment
SCADA:
Supervisory Control And Data Acquisition
SCM:
Supply Chain Management
SDS:
Simple Dynamic System
SIC:
Sensitivity to Initial Conditions
SIMD:
Single Instruction on Multiple Data
SISD:
Single Instruction on Single Data
SME:
Small and Medium Enterprise
SMED:
Single Minute Exchange of Die
SMI:
Small and Medium Industry
SPC:
Statistical Process Control
SPSM:
Self-Production System Monitoring
SPT:
Shortest Processing Time
SSPR:
Single-Step Production Reservation
TAT:
Turn Around Time
TCM:
Thermal Controlled Module
V&V:
Verification and Validation
VAC:
Value-Added Chain
VFDCS:
Virtual Factory, Distributed and Control System
VOD:
Video On Demand
WIP:
Work-In-Progress