Dear Dr. Rai,

Your question is very interesting indeed if we consider that in our days a large family of systems is developed which has some or multiple self-reflective capabilities such as self-organization, self-adaptation, self-evolution and self-reproduction. However, we have to be considerate. The term 'system' is simply too broad (practically covers everything). Are we talking about natural (physical, biological, social or transcendent) systems or engineered (technical or hybrid) systems? If the latter is in the focus, then are we talking about ordinary technical systems, complicated technical systems, or complex technical systems? Or complex socio-technical systems? To arrive at a meaningful answer to your questions, it is wise and advisable to decide first on what kind of system or kinds of systems are we interested in, since their properties and their interactions with other systems and the embedding environments are very different.

Kind regards,

Imre Horvath

I completely agree with Imre Horvath, who has furnished an insight to explain what are systems all about. However, I wish to venture one step beyond and randomize the whole problem. Shall I invent a new term for this problem scenario ... How about the concept of system longevity then? In that case, we may think about systems having lives. For example, the system of a protozoic cell division may have a life of a few microseconds. But the solar system has got a life of millions of years. What I want to say is that the issue is all about the survival capacity of the given system. Does a stable system survive more? I don't know of any case in the realm of physical and engineering sciences that a stable system cannot survive. A converse is that an unstable system tends to fall apart, transform, or cease to exist. So, if a system is stable, it is a presupposition that it would survive and live for long. As far as the environment is concerned, we can consider a system to be a set of things that exists inside the universal set or S. As such, the dynamics and conditions of system longevity will start compounding with respect to themselves and numerous variables will have to considered. A certain degree of unpredictability will also have to be borne in mind while calculating the survival possibilities of the system concerned.

Dear Veerendra,

I guess your question targets information systems. IS are very complex, and they integrate various aspects – IT, managerial, psychological, sociological, political … I believe that people would keep the power to manage information systems, but sometimes it seems that systems manage people...

So, your question is very interesting, and it is the IS developers’ dream – how to manage the constant changes needed in IS that are provoked by internal changes in the enterprise, or external changes in the environment. I.e. the IS should be adaptive and responsive. Moreover, the systems should be changed easily and quickly.

In my opinion,modeling techniques and business process management (BPM) technologies could help solve these problems in some cases. Cloud services and the BPaaS model are also very promising in this context.

Good luck,

N.F.

When trying to summarize this question with regard to enterprise architectures, I've selected four features: comprehensiveness, modularity, clarity of principle, and consistency.

https://caminao.wordpress.com/2015/10/15/eaf-non-nego-feats/

Dear Colleagues,

A nice discussion is forming ...

With systems, and systems development, we reached to a specific age that creates a new challenge not only because of the growing complexity (in other words, the tendency to show some form of emergent behaviour), but also due to the challenges posed by the increasingly tight interactions of systems with their embedding environments and stakeholders, as well as by the smartness-enabled self-* capabilities of the established technical and socio-technical systems. The largest challenge what I see originates in this very last issue. Not long ago, we tried to design (define, explore, document) our systems exhaustively from functional, architectural, behavioural and interactional points of view in the design stage. However, present time systems (such as cyber-physical systems, which my research team is focussing on), penetrate deeply into real life processes, collect large amount of content and context information through sensors, build awareness and reason in context based on statistical or/and semantic processing of these data, and actuate interventions into the physical processes in order to achieve some objectives, which are not necessarily defined completely in the design-time.

Non-linearly behaving systems are able to learn and adapt themselves run-time according to what they learnt and what their overall objectives imply. These run-time exercised self-resistance, self-organization, self-management, self-evolution, self-reproduction capabilities have not yet been sufficiently studied either from scientific and engineering points of view, or from philosophical and teleological points of view. The major design concerns related to the design-time handling of the run-time acquired dynamic behavioural characteristics are: (i) what positive capabilities can be and are obtained by certain systems, (ii) how the run-time obtained information and capabilities can be harmonised with the defined objectives of system operation, (iii) how large and what kind of deviation from the initially intended objectives can be accepted, (iv) how the deviations can be controlled and how new positive objectives can be considered by designers and the systems, (v) how large autonomy can be delegated to and accepted from the perspective of systems, stakeholders, environments and/or society? The incomplete knowledge entails uncertainties for complex system designers.

What is your opinion about these issues please? Kind regards,

Imre Horvath

Dear colleagues, Data Analytics is in vogue these days. It is a powerful tool. But, it has limitations. It can not accomplish everything we need. My intent of asking this question is to check if we need "Systems Analytics" also. What do I mean by "Systems Analytics"? I mean analyzing a socio-technical system comprehensively to see if the system has desired set of characteristics / properties? What are the desired properties to look for? How to look for those properties and how to embed those properties in the system when we design it? We need a debate around the issue of "Systems Analytics". Thank you!

Well, Mr. Rai, are you talking about socio-technical systems? In that case, the need of historiography cannot be overlooked. If we try to work with a new conception like "systems analytics" altogether, then we have to emphasize the evolutionary aspects of the system under examination. Best assessment of evolution is only possible when we look at the history of development of the system.

I tend to resonate with Dr. Ray. Social-technical systems have a dual-face as far as their innovation, development, implementation, operation and evolution are concerned. At the bottom, they are engineered systems, such as a rail transportation system, the infrastructure, vehicles, control, scheduling, services, informing, etc., of which are results of (multi-disciplinary) engineering work. On the other hand, they are also social systems, which are however driven by less controllable social factors, and reflect some historical/cultural evolution too. In the context of the above rail transportation system it means that, if the timetable is not relevant for the people or the coaches are always crowded with other travelers, then people will prefer other transportation opportunities. Obviously, this may soon be recognized by those operating the transportation system and they will most probably consider some sort of technical solution first (such as either a more favorable timetable, or trains with more coaches or seats). But social solution can also be considered since people may change their daily routines. What I would have liked to indicate with this example is that socio-technical systems should be considered and analyzed as ecosystems. Purely technological or purely social analytics alone will not lead to proper insights. Ecosystem analytics typically considers value propositions as a possible viability measure. (Other aspects may be dynamic balance or equilibrium, or engineering and social value co-creation).

TWEETSTYLE ANSWER!

Existing AI systems that set goals for themselves and create and execute plans to achieve them seem to me to be purposeful. Eg my "Pleasure Seeking Automaton" publication on RG. Recognition seems hard --- from behaviour or from observable internal properties?

Best wishes

Jim (+MoHAT)

Dear Jim, your valuable comment stimulates me to ask how much and in what ways existing AI systems can recognize, consider, deal with, and benefit from non-predefined social interactions and developmental influences in plan making? There might be some solution for this, but I am not aware of any sufficiently sophisticated ones for the cases when dynamic aggregative complexity and quasi-spontaneous emergence are at stake. Whereas, these two are among the primary paradigmatic features of real life socio-technical systems, as Kroes’ and others’ studies have discussed it. Regards, Imre

TWEETSTYLE --> Imre

I agree. I believe that current AI “purposive systems” [using my definition] are as yet nowhere near your target socio-technical systems. Though I do wonder what the LARS control software designers are thinking about for their next generation(s)...

Regards

Jim

Two alternatives:

1. Antifragility as defined by Taleb

2. Viability as defined by Beer.  VSM - Viable System Model gives you a general answer, though its implementation for a given environment being far from a simple matter. The structure that works for one environment might be be useless for another.

Cheers/Ilia

P.S. Here is a paper of ours that tries to figure out heuristics of what kind of business processes, technology, techniques and culture could be suitable for a particular business environment: 

Slides: http://slidesha.re/1qEVcz1

Paper: https://www.researchgate.net/publication/282074706_A_Framework_for_Synchronizing_Human_Behavior_Processes_and_Support_Systems_Using_a_Socio-technical_Approach

Conference Paper A Framework for Synchronizing Human Behavior, Processes and ...

tolerance .

I think you need flexibility and adaptability most of all.

I will say that it takes prospective and monitoring process coupled with the company's strategy with  different business cultures networks to exchange and make recommendations. Agility, openness, creativity, flexibility !

Hi Imre (others), You mentioned earlier ; "The largest challenge what I see originates in this very last issue. Not long ago, we tried to design (define, explore, document) our systems exhaustively from functional, architectural, behavioural and interactional points of view ...". 

I am trying to map complex mining processes (people and technology interactions) so that a simulator can be built. I am interested in the most effective method/language/ tool you found for representing the complexities, logic, interactions, prows, activities. I am particularly interested in getting (non-IT) knowledge-owners to describe and see so that IT people can see and code the system. BPMN and UML languages don't hep as they need translation for illiterate knowledge-workers and take too long to learn??

Thanks

Mark,

Halpin's Object-Role Modeling is the only formal, consistent, complete, and non specific language. (Ontology-based languages are specific or partial).

Halpin Terry. “Information Modeling and Relational Databases, 2nd Edition”, Morgan Kaufmann (2008).

Thanks Remy. I can see the appeal of ORM for data modelling but need to look deeper for business processes.

As a matter of fact business process modeling can be seen as adding two dimensions to data modeling, one for activities and flows, the other for states and transitions. Since ORM constructs are formally defined for instances they can be applied uniformly to static (data) or dynamic (activities and states) ones.

I am interested in your opinion:

Is there a difference, and if so, what is the difference between designing complex socio-economic systems.and designing complex socio-technical systems?

Can socio-economic systems exist without having technical components?

Technology is with human activities from the beginning, and economics with human exchanges also from the beginning. The difference is between social systems and mechanical ones, the latter can be fully designed, the former cannot.

I agree with #1, but #2 is problematic because it's a necessary condition but not a sufficient one with regard to purpose. For that you have to introduce consciousness, and therefore the distinction between design and emergence.

Dear Dr. Rai and Participants,

                                I have summarized my research on Physical and Econophysical Laws of Systems Engineering by using a stock market systems DGE model under uncertainty with microstructure heterogeneity and information technology which is most general and complex DGE situations (with production) , which is a quantum mechanical system with a diffusing political environment and systems generated endogenous genetic learning and have derived four laws viz. 1) Information Law of Demand 2) Financial Law of Demand 3) Systems Classification Theorem 4) Systems Integration Theorem, the four of which are Quantitative Finance Laws as much as Quantitative Quantum Mechanical Laws of Open Stock Markets and I have used String Theory + General Relativity + Fractal Geometry to Physically*Economically^Financially construct and embed these electronic and information highway stock exchanges in the countries capital (or financial policy centres). These Stock Markets reach equilibrium within natural planning periods and genetically grow the pool of knowledge available to investors besides the other functions, and are sustainable in the sense of integration with their environments. You can take a look at my papers on Haag Theorem and Stock Markets, Political Environment and Stock Markets and Bounded Rationality and Stock Markets on my RG site, if you would like to. SKM QC FEPS

https://www.researchgate.net/publication/305815882_Challenges_for_Atomix_System_Ltd_The_ITES_industry_in_Bangladesh

Article Challenges for Atomix System Ltd.: The ITES industry in Bangladesh