by Peter P. Yim, President, CIM Group of Companies

Presented at the ICCIM'91 Conference (Oct. 2-4, 1991, Singapore)


CIM3 is an approach to computer integrated manufacturing proposed by the author in 1990, which extends the traditional CIM architecture and design to incorporate the general paradigm of Computer Supported Cooperative Work (CSCW).

The CIM3 system models the manufacturing enterprise as "coordinated workflows" through a system of integrated automation and communications. The use of WORKFLOWs as an integrator between the HUMAN and existing CIM automation platform form the key approach to the CIM3 system design.


It is already the 18th year since the term "CIM : Computer Integrated Manufacturing" has been coined [Joseph Harrington, 1973]. The concept is by no means very popular yet. But then, it is already moving into its second decade of serious work (admittedly, not a lot was, and could have been done during the '70s due to cost and technology constraints).

"Have manufacturing companies or industries, as a whole, improved through the synergy gain from the past decade's integration efforts for those who opted to invest into Computer Integrated Manufacturing?"

The answer leans towards a "NO"; and that there are definite fundamental and strategic issues which need to be addressed during this decade as pointed out by Dr. Joseph Lau of the CIM Group in his paper "A New Look at CIM -- Strategic Issues for the '90s" [7].

Like the case of industrial engineering, where work and people has to be organized differently to obtain the productivity gain in mass production oriented factories (as opposed to the craftmen and the cottage industries), the computer integrated manufacturing companies need fundamental changes in terms of how work and people are structured and managed (like adopting the task/team concept or the flattened organization) before the real effect of all the computing power, connectivity and information sharing can be fully realized.

We can look back and see that a lot of the technical issues were sorted out during the '80s. Earlier, we could not even connect two machines from different manufacturers together reliably (or even have a set of standards to go by), without a major effort. Today, we can almost take Local Area Networking for granted.

Recognized by the computer industry in general, a lot of the issues we are left with, today, are human issues. Better user interface, groupware (computer software which aims at improving productivity for teams or small groups of people, rather than for individuals), etc., seem to be some of the answers that will emerge this decade.

Manufacturing (in the broad sense of the word -- which entails ALL activities of a manufacturing company -- encompassing strategic planning, marketing, product development, material procurement, logistics, production, distribution, management and coordination etc.) is a very complicated exercise, even to put it mildly. A lot of earlier attempts in computer integrated manufacturing has been aimed at building FULLY AUTOMATED manufacturing concerns which are operated by very few or even no people. Is the "lights out factory" really the appropriate metaphor for the "manufacturing company of the future"?

The author purports that, at least for the time being and probably for quite a few more years to come, the most critical success factor in manufacturing enterprises are its people, particularly, it management. This prompted him to embark on research on improving effectiveness of manufacturing companies ever since he founded the CIM companies in 1984. In early 1990, the author coined the term "CIM3" as the acronym for "Computer Integrated Man-Machine Manufacturing" to paraphrase his concept of extending computer integrated manufacturing from a hardware-based (manufacturing equipment, computers, networks etc.) system model, or application software-based (information system) model, to that of an integrated man-machine system, where the human is as much a part of the system as the hardware and software that are involved. The objective is to come up with the system model, the system design and implementation methodologies, as well as the organization and structure that would optimize the effectiveness of manufacturing concerns. This paper shall highlight some of the major concepts developed or applied.


The deficiencies of most existing computer systems to cope with the irregularities of the real world has been recognized in the recent development of an emerging research area called COMPUTER-SUPPORTED COOPERATIVE WORK ("CSCW") which takes as its primary focus the roles in which computer systems play in the organization of joint effort.

In CIM3, the author attempts to extend the traditional computer integrated manufacturing architecture and design to incorporate the general paradigm of CSCW. One key concept in CSCW is that of the WORKFLOW. Workflows are the patterns by which activities and communications are being handled by human beings in real life [11]. CIM3 employs WORKFLOWs as a super-layer in the system architecture to integrate PEOPLE with the Computer Integrated Manufacturing automation platform (as currently perceived) to form INTEGRATED MAN-MACHINE SYSTEMS.

Rather similar to the vision of Dr. Robert Johansen of the Institute for the Future in his vision of the "orchestrated workflow" as the enterprise model for the future [6], the CIM3 system employs WORKFLOWs as the MAN-MACHINE INTEGRATOR to arrive at its system model. Under this system model, WORKFLOWs provide the integration between the human, the management and the people aspect of manufacturing on the one hand, to the automated equipment, connectivity and information system aspects of manufacturing on the other. With its human element in the system, CIM3 will be a "people oriented" , "more flexible" and "result oriented" system.

CIM3 shall also provide the framework whereby human intelligence plays a role when needed, but at the same time, allows it to be substituted by artificial intelligence when the technology of the latter is mature enough to take over.


The underlying theory behind CIM3 is in the Language/Action Perspective to Systems Design, and how to Coordinate them, as proposed by Dr. Fernando Flores and Stanford University's Professor Terry Winograd [4, 5 & 11] which calls for a shift from a "Planning/Knowledge Perspective" or an "Information/ Communication Perspective" to view activities in the enterprise as people carrying out "Conversations" which forms the basis for all "Actions" within the enterprise. The term and the notion of "Conversation" has, more recently, been replaced by that of "Workflow".

Action Technologies, Inc. ("ATI", a Company founded by Flores, Winograd and others) applied the Language/Action paradigm in their "Business Design Technology" as elaborated by Robert Dunham [3]. They have proposed a methodology in analyzing the workflows in the OFFICE and the BUSINESS enterprise, and expanded on the seven fundamental patterns in CONVERSATIONs (namely those which open with REQUEST, OFFER, PROMISE, INFORM, QUESTION, WHAT-IF and NOTE). They proposed a system for graphically representing the flows and states of workflows (which this author shall adopt), a basic example of which (that which opens with a REQUEST) is shown in Figure 1.

In terms of system analysis and design, CIM3 extends from the above works in the context of the manufacturing enterprise. Basic workflows for manufacturing will be identified. The inter-relationship between these workflows and other resources in the manufacturing enterprise (e.g. machinery, engineering personnel, etc.) also needs to be properly defined in a System Architecture.

A fundamental question shall be asked : "What is the objective in going to CIM3?" As we have identified CIM3 as the integrated man-machine system which models the manufacturing system in question, IT IS, in fact, THE MANUACTURING ENTERPRISE model. As such, its objective should be IDENTICAL to the objective of the enterprise. Along the line of thinking of ATI & Dunham [3], the objective should be one of CREATING CUSTOMER SATISFACTION. Whether these "customers" are external or internal, as in the enterprise's shareholders, clients, suppliers, bankers, or staff, workers, managers etc., the objective for CIM3 shall be to allow these people, information system and process equipment etc. to work together MORE EFFECTIVELY ("coordinated") towards achieving the goals and objectives of the enterprise.


While the lower layers of CIM3 shall comprise of the more traditional CIM approach of system integration through equipment connectivity, information sharing and integrated databases, these are supposed to be further integrated (or at least, interfaced to) the layer of WORKFLOWs comprising of human interactions through structured and/or even free-form language communications.

Figure 2 depicts the CIM3 System Architecture.

The fact that the uppermost layer of human users (managers, engineers, developers etc.) may interact among themselves in human languages (which are NOT pre-programmed), we are looking at a system that is truly intelligent (as opposed to being of an Artificial Intelligent one), more flexible, and has the capabilities of handling exceptions as well as spontaneous activities.


Extending from the Product & Order Cycle categorization by Professor Marvin Manheim of Northwestern University [8], we shall model the Manufacturing System as Workflows within the following four major Cycles of the typical Manufacturing Enterprise :

While the above workflows are generic in nature, they cover the bulk of the activities and workflow vocabulary of the typical manufacturing enterprise. For specific manufacturing operations, certain augmentation may be necessary so that all activities within the enterprise may be encompassed.

The Material/Information/Financial dimensions of manufacturing flow [1] may be analysed by taking the Language/Action Perspective and modelled in terms of workflows as in the following examples.

(a) The conversion of raw materials into finished products shall be viewed as the workflow which begins with a client requesting products (through the placement of a Customer Purchase Order) and ending which that client declaring satisfaction (by indicating that the products have been received in good order via, say, the issue of their Cargo Receipt and Incoming QC Acceptance Report).

(b) The flow of information and data shall be viewed from the perspective of, say, management requesting reports and ending with the completion by the system (the integrated man-machine system in this case) to provide those reports to management satisfaction.

(c) Likewise, the Financial flow shall be modeled as the request for funding to support, say, a product development project or a work order (prompted by having received a customer order).


As shown in the System Architecture diagram before, a key design approach will be in the use of Object Oriented Design.

Three key OBJECT CLASSES have been identified to encapsulate all workflow activities. They are :

Analyzing and designing the CIM3 System can thus be accomplished through a combination of Object Oriented Analysis and Design [ref 2], Structured Analysis and Design Technique [ref 9], Augmented SADT Technique [ref 10] and Business Design Technology methodologies [ref 3]. The development of an integrated CASE (Computer Aided Software Engineering) methodology and its tools employing the abovementioned techniques is being proposed and will not be elaborated in this paper.


CIM3 system is one which is "people" oriented and "business objectives" oriented. It is consistent with TOTAL QUALITY Management, as well as being a "pull" system that is consistent with the "Just-in-time" concept. CIM3 is envisioned to be truly intelligent, more flexible and responsive, and has the ability to handle rules as well as exceptions, planned as well as spontaneous activities and still provide the necessary management control and traceability.

To differentiate CIM3 from CIM, we should note that :

Departing from the direction towards the "dark factory" and aiming towards integrating "people" (and especially management personnel) to form truly intelligent integrated man-machine manufacturing systems, the CIM3 approach provides solutions to some of the pitfalls of the '80s styled CIM system concepts and promises to address most of the issues posed by the integrated manufacturing enterprise of the '90s and beyond.


[1] Baudin, M. "Manufacturing System Analysis" Yourdon Press; 1990

[2] Coad, P. & Yourdon, E. "Object Oriented Analysis" Yourdon Press 1990

[3] Dunham, R. "Business Design Technology : Software Development for Customer Satisfaction" Proceedings of the 24th Hawaii International Conference on System Sciences; January 1991

[4] Flores, F. "Management and Communication in the Office of the Future" Doctoral Dissertation, University of California at Berkeley, 1981

[5] Flores, F. & Winograd, T. "Understanding Computers and Cognition: A New Foundation for Design" Norwood,NJ; Ablex, 1986

[6] Johansen, R. "Teams for the Future" Proceedings of the 24th Hawaii International Conference on System Sciences; January 1991

[7] Lau, J. "A New Look at CIM -- Strategic Issues for the '90s" Proceeding of ICCIM'91; October 1991

[8] Manheim, M. L. "Global Information Technology : Issues and Opportunities" Proceedings of the 24th Hawaii International Conference on System Sciences; January 1991

[9] Marca, D. & McGowen, C. "SADT : Structured Analysis and Design Technique" McGraw-Hill; 1988

[10] Marca, D. "Augmenting SADT to Develop Computer Support for Coorperative Work" Proceedings of the 13th IEEE International Conference on Software Engineering; May 1991

[11] Winograd, Terry "A Language/Action Perspective On the Design of Cooperative Work" Proceedings of the First Conference on Computer Supported Cooperative Work; 1986

[12] Winograd, T., Newman, N. & Yim, P. "Including PEOPLE in Computer Integrated Manufacturing Designs" Proceedings of ICCIM'91; October 1991

[13] Yim,P. "System Integration for Industrial Automation" Industrial Engineering Journal, IIE(HK); 1986/87

[14] Yim,P. "CIM -- From Product Design to Manufacture to Distribution" Proceedings of the IA'90 Conference on Industrial Automation, Singapore; May 1990

[15] Yim, P. "CIM3 : A Vision for the Next Generation of CIM" Proceedings of the Conference on Technology Management Strategies for the Global Enterprise; CWRU, Cleveland, OH; May 1991.

-- end --