Center for Open Enterprise Management


Executive Summary

The Center for OEM is a proposed cross-industry research center at Stanford University, sponsored by major companies with a strong presence in the Silion Valley such as CISCO, SAP, and at least six others to be announced. This NSF-sponsored center has the mission to develop a radical solution to two emerging serious problems in the IT industry caused by the increase in complexity inside and among enterprises. One problem is that hand-coding of processes will not scale. The other is the lack of support for coordinating the activities of individuals in different departments and companies: "Collective Work".

We propose to develop a radically different approach to solving these problems based upon compuational logic and semantics, and the analogy of "enterprise physics" in which the laws and policies of the enterprise(s) are expressed and executed directly without the interpretation and mediation of conventional programming. A variety of new technologies are envisioned including new ERP systems and "smart routers".

Further, Stanford proposes to develop jointly with our industrial partners "The Digital Department": a prototype actually used to manage the complex operations of the Stanford Computer Science Department. This will not only be an existence proof of a revolution in IT, but will be a showpiece for the Center for OEM and each of its industrial affiliates.

Introduction

Enterprise Management is the process of managing the internal resources of an enterprise and its interactions with other enterprises so as to maximize its goals while complying with applicable regulations and contracts. Specializations of enterprise management include resource planning, human resource management, customer relationship management, supply chain management, and so forth.

Over the years, enterprises, large and small, have discovered the value of computer support for managing such matters, leading to Digital Enterprise Management(DEM) in which much critical information is expressed digitally and can be manipulated by Enterprise Resource Programs and other enterprise management systems. Many companies have grown and flourished in this space, often described as ERP systems.

Examples of software vendors include HP, IBM, Intuit, Oracle, Peoplesoft, SAP, and others. Examples of consulting firms include Accenture, and Siebel. CISCO has developed sophisticated routers to support the enormous growth in information transfer due to interactions among departments within enterprises and among enterprises. The success of DEM has led to a severe problem.

As companies become more complex and interact with each other over the Internet in value-added networks, the current paradigm of having programmers implement the corporate policies and their intra- and interconnections is not scaling. Managing change and verifying conformance is adding unsupportable costs to enterprise balance sheets, which will ultimately lead to the case of the IT tail wagging the business dog.

There is a further problem that is less-recognized but increasingly important as companies become more complex and interconnected. Individuals in different departments and even different companies must work together in order to achieve their individual objectives, often without knowing each other. There is today no computer support for the coordination of these people, which we describe as "Collective Work".

Especially, there is no automatic notification that one person's actions might help or conflict with the objectives of another. Nor is there any way to make requests of people whom we do not know except by the their abstract qualities: whoever is in charge of procurement or of the conference rooms, or who may be working with our Japanese subsidiary. Our people networks today are based solely on who we know personally and what we know about them, which limits the power of individuals to achieve their enterprise objectives.

Open Enterprise Management

Our approach to solving this problem is called Open Enterprise Management (OEM): allow policies to be expressed directly and declaratively as "laws of the enterprise" and also provide a framework that ensures that these laws are respected and applied as needed in order to manage the enterprise, including all processes, regulatory compliance, and supply chain management. This effectively provides a virtual physical universe in which "forces of nature", rather than hand coding, makes things happen, and changes to policies are implemented just by changing the policies. The result of such "enterprise physics" is: support for all parties in enterprises - individuals, companies, and governments, and lowered cost for more efficient interactions.

Our approach is based on a formal model of the enterprise captured in computational logic. All state is captured in databases and associated constraints that conform to this model. Definitions, policies, regulations, and contracts encoded as logical sentences (more expressive than ordinary business rules): these are usually referred to as "semantics", though the Stanford model is simpler and more pragmatic than usually espoused in academic conference papers. This includes semantic web services in which services are also described sufficiently to be used within the context of enterprise laws.

The objectives of the company and individuals are achieved, while respecting the laws of the enterprise, by an overarching framework for finding the logical consequences of these rules (versus mere firing of condition/action rules.) A unique feature of Stanford's computational logic is the detection of inconsistencies and the ability to continue reasoning even while they are yet unresolved.

Further, we use these capabilities in unique ways in order to computationally support and facilitate Collective Work: distributed people pursuing multiple objectives, the partial achievements of which conflict and reinforce in ways that should be known as soon as possible. Some of the approaches being developedre:

This is all based upon an underlying semantic model of all data, people, and services within the enterprise. This in turn allows individuals to maintain personal descriptions of themselves and the services they own, which are then used by the entire enterprise. Individuals are empowered not only in personal information management and coordination with each other, but also in creating "on the fly" mashups of all enterprise services to achieve their work goals. Special tracking and notification systems will be developed after study and analysis, similar to those previously developed for concurrent engineering.

Supporting the enterprise will be "smart routers" that will use semantics to efficiently connect services and people, routing the information based upon content and ensuring that every service and person recieves the requests, answers, and notifications necessary to ensure the smooth and efficient operation of the enterprise.

IT then concentrates on maintaining the underlying computational framework and the semantics of data integration, building upon prior expertiese in database management.

The Stanford Center for OEM

Stanford will take advantage of an NSF grant in order to leverage the research sponsorship of industry. The target for this grant is to have eight (8) affiliates that each contribute $50K each. Each company becomes an Affiliate of an official Stanford research center, witht the usual rights and privileges. IP terms will be determined by the NSF and apply equally to all Affilitates.

The staff of the center will include a core of full-time employees, together with faculty and students from Stanford and professionals from affiliated organizations. Industry will be encouraged to have visiting researchers. The center will by hosted at the Stanford Logic Group and directed by Prof. Michael Genesereth. We anticipate furthering connections with the Stanford Business and Law Schools.

The center will build a prototype of the new enterprise management system: "The Digital Department". We build a new system for operating the Stanford Computer Science Department, which itself has various staff functions as well as professors and students, all of which have different objectives. In addition, the department must coordinate with and conform to the regulations of various Stanford and outside government agencies.

The Digital Department will be a prototype that will demonstrate the value of the new technologies. Furthermore, this is a large project that could not be achieved by a single group or even a single industrial sponsor. Each Affiliate will have a place in this technical ecosystem, and all will be required to work together in order to achieve this massive re-thinking of IT. Indeed, the first small prototype of OEM with be the operation of the center itself.

The Center for OEM will be a unique neutral setting at Stanford University for Silicon Valley IT companies to jointly develop new IT technologies for the 21st century, using the Stanford CS department as a proving ground and showpiece.