The Rise of The Intelligent Enterprise

Kemal A. Delic - Umeshwar Dayal

Web design and development by:

Page 1

The long history of fascination with human-like artifacts (machines) able to emulate certain behaviors dates back a couple of centuries when some famous people invented the whole range of automata being able to play music, dance, understand and write certain words or even appear "intelligent" [Living Dolls: A Magical History of the Quest for Mechanical Life - Gaby Wood's book].

More recently, in the 20th century, the mathematician Turing established a test for "artificial intelligence" (AI) according to which the ultimate proof is the inability of a human jury to distinguish answers from the machine from human answers. This eventually triggered half a century of research into AI.  Whilst the elusive objective of this research was to create AI itself, the most concrete outcomes are to be found at  lower and more modest levels in the same way that medieval alchemists trying to turn dirt into gold, while never reaching their original goal, nevertheless invented many valuable chemical processes and useful devices.

Thus, we can see today some AI-based, very practical solutions embedded, for instance, in modern cars (anti-skating), medicinal devices (initial image analysis), industry (robots, voice-recognition, scheduling of resources) etc. It would seem that AI is presently most successful where it is somewhat hidden in the form of crucial, but dependable and working embedded technologies. One may argue that developments in the field of Information Technology have been the key for such "intelligent developments".

Science fiction (SF) authors and movie makers have had much more freedom to indulge their dreams and fantasies in this area. Being futurists and projecting possible world(s) of tomorrow, they have depicted intelligent things, humanoid robots, smart cars, intelligent buildings and cities. While such visions have not been entirely fulfilled, some correspondence with modern reality can be discerned. In a similar spirit, this article will outline a vision of the future "intelligent enterprise" (IE) - a business corporation morphing into complex system able to behave as the biological system or ultimately to act as the intelligent human.

Intelligent Enterprise Defined

The notion of the IE is characterized by the ability of businesses to morph into new forms/entities with some surprising results. Today's enterprises deploy IT to do their business, which is introducing significant complexity.  This technical complexity combined with organizational complexity can and does create internal "islands of inefficiency" which are difficult to deal with. We envision that in the future "Intelligent Enterprises" will be able to (automatically?) transform themselves into better forms by becoming "more intelligent".  They will derive efficiencies through the automation of their core business processes, and the exploitation of knowledge. Agility will be improved by one or two orders of magnitude, intelligence density within the enterprise will grow sharply and problem-solving capacities will be dramatically enhanced. They will form dynamic partnerships with other enterprises to create dynamic business ecosystems, which will be self-managed, self-configured and self-optimized. Consequently, they will exploit new forms of business value creation, adjusting their behaviors to the markets by learning from them.

The question may arise why we are comparing engineered systems to biological, natural systems ? And why we assume that the adaptation is a key behavior, among all others ? This came from our belief that both systems share the same ultimate objective : to survive in evolving environment and changing circumstances.  To achieve this, they should basically be able to sense it's environment, to understand situation and to create viable plan which will be then reliably executed. This seems as the situation in which both systems exhibit also learning behavior. It is very likely that engineered systems will never reach the level of sophistication, elegance and beauty of the nature-born systems, while from the pragmatic point of view, they will serve as the ultimate ideal to strive for.

Looking at  current large and successful businesses, one can observe that many have already deployed some primordial forms of the future intelligent enterprise. The current acronym soup of CRM, ERP, EAI, EKM encompass enterprise activities aimed at improving efficiency or injecting intelligence into their operations.

Figure 1 : Generic Intelligent Enterprise Architecture (Click here to view full size version)

CRM (Customer Relationship Management) serves to understand, maintain and improve evolving customer needs. ERP (Enterprise Resource Planning) addresses issues of supply chain efficiency and back-office optimization. EAI (Enterprise Application Integration) aims at the integration across islands of functionality, by not only providing connectivity and communication, but through the orchestrated functioning of all enterprise sub-systems. EKM (Enterprise Knowledge Management) is a broad area whose ultimate objective is to inject knowledge into business processes, enable reuse of human expertise and to create system artifacts with rudimentary intelligence.

Intelligent Enterprise is therefore:

1 - agile - since it is able to launch production order very rapidly after, for instance, a web-based purchase was closed

2 - adaptive, self-regulating, self-optimizing - since it is able to adjust the key business parameters (revenue, profit, cost) to the short-term, changing business climate 

3 - self-defensive - as biological organisms reacting on various attacks with appropriate biology-inspired resistance mechanisms

4 - with fuzzy borders, mesh-like structured - since it is able to restructure and scale the organization to fit dynamic needs. As in outsourcing or complete outsourcing, for example  

5 - self-aware, aware of the markets and able to learn from them and adapt to them - since it is able to deal with internal inefficiencies, bottlenecks and latencies and to coordinate them with perceived market changes. As knowing, for example, the level of goods being produced,  able to correlate this with the market niche real-time situation and jump quickly into price reduction campaign on the Web

6 - able to morph into new and better forms - this is an long-term process in which we may not even guess what will the final form look like - it will include not only adjustments of the business and operational parameters, but also major structural changes : deciding for mega-merge or drastic resizing while changing the industry branch, for example

In a very rough analogy shown in the Figure 1, we may recognize enterprise sensors, effectors, nervous system and brain corresponding to characteristic constituents of living organisms, each with particular functionality. CRM provides sensing functionality; EAI may represent nerves interconnecting ERP effectors with CRM and EKM being dispersed across enterprise, but centrally coordinated from the brain (company top management).The best current guess is that the future IE will be hierarchically organized, decentralized and distributed system with high level of self-regulating, local autonomy. Some authors are reporting about implementation of such systems combining component-object technology with intelligent agents technology to create complex adaptive systems [1], using web-services as the glue for heterogeneous enterprise applications [2] or a wide variety of enterprise middleware technologies for integration purposes [3].