This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License. Industrial Operations / Information Processing Convergence Control Chain Management Body Of Knowledge MI Science for Enterprise Systems 04/2011 Jean Vieille www.syntropicfactory.com j.vieille@syntropicfactory.com Research community www.controlchainmanagement.org Consulting group: www.controlchaingroup.com Agenda Enterprise as a system Entropy Chaos Complexity Information Linguistics MI - Science for Enterprise Systems 2 Systems theory studies Open, Complex systems System theory is not exactly a science Ø "A set of theories - automata theory, linear systems theory, control theory, network theory, general Lagrangian dynamics, etc. - unified by a philosophical framework" Ø a holistic approach to reality System: Ø Set of elements in dynamic interrelation that are organised for a given purpose (J. De Rosnay) Open system: Interacts with its environment Ø =/ Closed system: no matter/energy I/Os Feedback loops on internal and external variables (Cybernetics) Complex system: the whole more than its parts, chaos... Opposes to (or complements) Cartesian, analytic approach Particularly applied in sociology, biology and environment MI - Science for Enterprise Systems 3 Types of systems Biological systems use membranes Ø (Cell, tissues, body) to insulate components and handle interactions. Ø They keep their structure, but renew at the cell level et reproduce Ø Slow, supra-genetic (Darwinian) evolution Artificial systems composition is generally variable Ø Social, mechanical, IT Ø Their structure can evolve rapidly in reaction to internal and external pressures Ø Meta models of these systems evolve like biological systems § Democracy, monetary system, transportation, computers, telephone Ø Shared components § People can be parts of several organizations § They can be employees and shareholders in the same company MI - Science for Enterprise Systems 4 The 10 commandments of Systems control Preserve variety Do not "open" regulatory loops Look for the points of amplification Re-establish equilibriums through decentralization Know how to maintain constraints Differentiate to integrate better To evolve, allow aggression Prefer objectives to detailed programming Know how to use operating energy Respect response times (J. De Rosnay) MI - Science for Enterprise Systems 5 A cosmic view of the World We know nothing about where we come from, where we go Ø More we learn about Nature, more we become conscious of our understanding shortness Different possible scales of the Unniverse Ø The human Era § from great apes evolution to final (eg. nuclear) self destruction Ø From Big Bang to Apocalypse § From energy soup to pure spirit Ø Part on an infinite cycle § Big bang succeeded to a previous cyle Ø Multidimensional § Other +/- parallel Worlds We evolve in a very narrow space-time spot! Ø Let's build our own little story about industrial enterprises... MI - Science for Enterprise Systems 6 Enterprise is an Open, Complex System Physical interactions: Ø Earth, other Enterprises, internal Resources Noospherical interactions: Ø Goals of the World, Humanity, Humans, Owners, other enterprises Social interactions: Ø Nations, NGOs, Trade unions, Family A thermodynamic entity Ø Consumes energy, applies it on matter Ø "Heat" (waste energy) when badly controlled (thermal entropy) Industry is a major component in the Earth eco-system Fast rise of enterprises « Social Responsibility » concern Ø Cares about social, environmental and economical footprints Ø Various motivations MI - Science for Enterprise Systems 7 Industry ecosystem (anthropocentric!) Agriculture Food Industry Biological-natural machinery Food Work Maintenance of Social Organization Industry anium Energy Industry Usable Oil, Gas, Nuclear, Wind, Sun, hydraulic energy Mechanical-artificial machinery Inspired by J. de Rosnay, Le Macroscope MI - Science for Enterprise Systems 8 Relevant sciences for Enterprise systems An enterprise is a complex, dynamic system As such, it is well covered by science and philosophy Ø applying for its understanding and control Though the applicable knowledge is rather unlimited, we will concentrate on some significant topics Ø seemingly helping best in our search of supporting the enterprise as an evolving organism within a harsh, constraining environment Ø This is an ongoing discovery work to highlight and leverage existing knowledge for improving enterprise sutainability This study concerns only the system aspects of an enterprise Ø Does not address specific business related mechanics § Physical transformations § Marketing and sales theories § Scheduling and financial management technics § ... MI - Science for Enterprise Systems 9 Enterprise as a scientific subject Largely studied subject Ø Ø Academic studies often stays at the "Valid philosophy" stage Intuitive, simple, common sense approaches (6 Sigma, Lean management, Theory of Constraints) more successful Constraints from Market, Shareholders, bankers, environmentalists, global economics => genial ­ or lucky ­ intuitive managers? Help achieving short term "reasonable" objectives & long term sustainable evolution Converge to Information Enterprises intuitive management Ø Ø Leveraging relevant physics principles Ø Ø MI - Science for Enterprise Systems 10 Agenda Enterprise as a system Entropy Chaos Complexity Information Linguistics MI - Science for Enterprise Systems 11 Entropy 3 meanings Ø Irreversibility: the 2nd law of Thermodynamics Ø Measure of the disorder: Kid's room, engineer desk... Ø Measure of ignorance: We are part of the system: Disorder prevents understanding Entropy of an open system can increase or decrease Ø Increasing = decrease of order, information Ø Decreasing = increase of order, information Can entropy be negative? MI - Science for Enterprise Systems 12 Entropy 1: Irreversibility (entropy macro observation) The 2nd law of Thermodynamics - Sadi Carnot (1824) Ø Based on observation of heat engines Ø over time, differences in temperature, pressure, and density tend to even out in a physical system that is isolated from the outside world. Ø Entropy is a measure of how far along this evening-out process has progressed Interpretations Ø In a system, a process that occurs will tend to increase the total entropy of the universe Ø Heat generally cannot spontaneously flow from a material at lower temperature to a material at higher temperature. Ø It is impossible to convert heat completely into work in a cyclic process - Engines produce unrecoverable heat Ø The Arrow of Time: Closed systems entropy always increases as the Universe's MI - Science for Enterprise Systems 13 Entropy 2: Disorder (generalized entropy definition) Fundamental definition Ø Entropy = the number of the possible microscopic configurations of the system Ø Entropy is maximal when all microstates are equaly likely Boltzmann (1896) Ø Statistical mechanics Entropy S function of W? Ø S = k log W Schrödinger (1944) Ø Linked entropy S to « state of disporder » D or « order » Or Ø S = k log D => -S = k log Or MI - Science for Enterprise Systems 14 Entropy 3: Information Hartley (1928) Ø « Quantity of information » H of a message of N signs in an alphabet of S signs Ø H = N log S Shannon (1948) Ø "Information Entropy" The minimum length of a message for a given meaning Ø Information inversely proportional to probability § Affected by coding, noise, redundancy § The entropy of a text in english is 1,0 ­ 1,5 bit/letter § More information = less probability, more complexity, more "chaos", more entropy MI - Science for Enterprise Systems 15 Entropy 3: Information Wiener (1948), Brillouin (1951) Ø The "opposite" Shannon' theory Ø amount of information = measure of its degree of organization, Ø entropy of a system = measure of its degree of disorganization Ø Information = Negative entropy (Wiener) = Negentropy (Brillouin) MI - Science for Enterprise Systems 16 Entropy 3: Information Stonier (1997) Ø The relation between entropy and information is not a direct linear relationship Ø From Boltzmann and Schrödinger § (1) S = k log Or Ø Direct relationship between information I and organization (or) § (2) I = c (Or) c to be defined Ø From (1) and (2) § (3) I = c e-S/k Ø "c" appears to be the information content I0 of a system when absolute entropy S= 0 § (4) I = (I0) e-S/k § Compare to the direct entropy / information relationships I = -aS Ø One entropy unit 1J/K = 1023 bits MI - Science for Enterprise Systems 17 Information / Entropy relationship Theillard de Chardin S = -k, I = ce -k « Spiritual » Universe S<0 I>c I S = Entropy = k log c/I I = Information = ce-S/k k = Boltzmann constant c = constant = Information at Zero S « Material » Universe limite S=0 ce I = c T= 0°K Current Universe state S>0 I