Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.stsci.edu/~rdouglas/publications/thesis/section2_4_3.html
Дата изменения: Tue May 13 19:35:11 1997
Дата индексирования: Sat Dec 22 19:13:18 2007
Кодировка:
Concurrent Engineering



Next: SNEAKERS Up: Introduction Previous: Goals of the

Concurrent Engineering

~ ~ Concurrent Engineering (CE) is a comparatively new design methodology, which enhances productivity and leads to better overall designs. CE has been identified as a vital ingredient in America's attempts to modernize its design and manufacturing practices. In the development of a product, there are many ``downstream'' aspects to be considered. These include final cost, manufacturability, safety, packaging, and recyclability. These aspects represent different phases in the product's life-cycle. In traditional design methodologies, the product is evaluated after each phase is complete. However, the downstream aspects are affected by decisions made during the design phase. Consequently, these aspects should be taken into account during the design phase.

In the CE scheme, these aspects positively affect the design decisions during the design phase. A team, composed of experts on each aspect, is brought together to participate in the design. These people have information about how downstream issues are affected by design decisions. Having information about downstream issues at design time has several advantages. First, having all of this information minimizes the possibility of needing to redesign some or all of the product. Eliminating redesigns cuts product development time and cost. Next, decisions which take advantage of particular features of an aspect can be made, such as choosing a set of parts in the design for which the manufacturing equipment is already tooled. Stoll [1986] gives an overview of these issues as they apply to the manufacturing aspect of product development. The practice of considering manufacturing needs at design time is known as Design For Manufacturing (DFM). CE attempts to extend the DFM principle to other aspects of the product's life-cycle.

Another advantage of CE is that, while knowledge is being built up about the design of the product, additional knowledge is being acquired about the other aspects of its life-cycle. As the design progresses, the manufacturing expert will learn more about how to manufacture the product, and the packaging expert will know more about how to package it, etc. This accumulation of knowledge helps to speed the product through the development process and get it to the customer more quickly, i.e., time-to-market is reduced.

Ingredients of Concurrent Engineering

~ ~ Based on research, twelve major issues have been identified in developing tools to support CE. These are the ingredients which have been found to be present in some form or other in many, if not all, of the systems studied.

  1. Design Agents are the principal actors in developing a design. They make decisions, monitor progress, or analyze parts of the design. These agents can be humans, expert systems, or other computer tools. They interact in order to develop a design.
  2. Multi-disciplinary Goal and Specification Representation is a means of expressing goals and design specifications in terms used by different disciplines, even though they use different terminologies. It is an attempt to unify the terminologies used by the Design Agents within the scope of a particular product design.
  3. A Catalog contains descriptions of parts that can be used in the design. These parts can be components or partial designs. They are a means of relating past experience to the current design, and speeding the design process.
  4. A Design Database contains the current representation of the object being designed. This is used by the agents when designing.
  5. An Accumulated Database contains knowledge about downstream aspects accumulated along with the design. This accumulated knowledge is discussed in Section below.
  6. Shareability of Design Information allows all Design Agents to use all design information, as they see fit. No information should remain the sole property of any agent, as there may be knowledge about a different aspect hidden in that information.
  7. Communication allows the agents to communicate despite their different backgrounds. Support for this may come through the development of effective communication channels.
  8. A Manager schedules agents, oversees negotiation, keeps track of resources used, and monitors and evaluates the progress of the design.
  9. A Design History captures knowledge about alternatives considered and decisions made, and the rationale behind those decisions. This is useful in determining if the best design was chosen, for credit/blame assignment, for learning, and for use in restoring the design if a design decision has to be retracted.
  10. A Checklist indicates important decisions that must still be made.
  11. A Standard Interface to all tools keeps the user from having to learn too many different interfaces. This lends a feeling of working in a single environment, and the addition of new tools into the system does not require a long learning period.
  12. Virtual Collocation of People makes all the agents appear to be in the same room. This promotes unity among the team members and encourages cooperation in determining a next course of action. It also encourages communication.
This list of issues was developed primarily from [Bedworth et al 1991], [Cunningham &Dixon 1988], [Jagannathan et al 1991], [Kott at al 1991], [Kroll et al 1988], [Lemke &Fischer 1990], [Lemon et al 1990], [Londoño et al 1989], [Stoll 1986], and [Subramanian 1990]

Views of Concurrent Engineering

~ ~ Clearly, as CE systems require many ingredients, many views of what is ``key'' are possible. Most of these views concentrate on the needs of the agents in the system (human or computerized) to communicate, or to have the right information available. This is one of the primary concerns of the Concurrent Engineering Research Center (CERC) at West Virginia University. They have put their efforts into indexed databases and electronic conferencing, in order to increase the amount of information that can be communicated to the necessary people involved in a design. (See [Jagannathan et al 1991]).

In a somewhat different view, Douglas &Brown [1992] suggest that the key issue in CE is the accumulation of knowledge. Thus the focus of this view is on ``what'' is being decided or learned, as opposed to ``who'' is deciding or ``how'' it is being done. The primary purpose of CE is to produce a design. That is clearly a process of knowledge accumulation from requirements to a design.

In addition, a CE system should produce descriptions of all other aspects of the manufacturing process, the assembly process, the design for packaging, etc. Thus the goal is to accumulate knowledge about all of the aspects of the life-cycle. Consequently, this view of CE as Knowledge Accumulation can be seen to underlie all CE activities, and is independent of the processes used to generate the knowledge, and the strategies for controlling them.

Figure presents a diagram that characterizes this knowledge accumulation process. Knowledge is accumulated during the design process about all relevant aspects of the life-cycle. In general, over time this knowledge moves from abstract to concrete, although in fact this transformation is more complex [Brown 1992b]. The design is complete when all relevant knowledge has been accumulated, and not merely when the component or product design is complete.



Next: SNEAKERS Up: Introduction Previous: Goals of the


rdouglas@stsci.edu