Products reflect more than just their components. They reflect the cumulative design, engineering, and manufacturing knowledge of the organizations that create them. Computer-Aided Design (CAD) supported by data management software captures and manages the "what" of products—the components. Knowledge-based systems, on the other hand, capture and manage the "why" and "how"—the cumulative knowledge. "Why" might include the reasons for selecting specific component features or materials. "How" might address the manufacturing process employed to produce a feature, or a description of treating a material to achieve the desired properties. The proper systems can optimally apply experience and knowledge to minimize the time spent on problem solving during the production of new products.

Well-structured knowledge management (KM) and knowledge-based engineering systems (KBESs) allow organizations to capture and reuse product development experience at many levels. They achieve this objective by progressively automating lower level repeated tasks, as well as higher level product development processes. To date, relatively few organizations have capitalized on this opportunity. Those who have report compression of the time and cost of design tasks by as much as 90 percent.

Despite this huge opportunity, companies ostensibly labeled as KBES suppliers have not yet experienced rates of growth comparable to other sectors of the software supplier community. The most common reasons cited for an organization's unwillingness to invest include:

• Management does not feel confident that making the major investment will pay off.
• The company struggles with the resource allocation required to support a knowledge management effort.
• The collation of experience and expertise—that is, best practices, methodologies, and decision rational—into a knowledge-based system proves to be complex and tedious within each expert domain and across disciplines.
• Operations personnel view knowledge-based systems as a risk to job security. Personnel may fear for their jobs when faced with knowledge-based systems that can perform tasks in a matter of minutes or hours that previously took weeks or months.
• Knowledge-based systems built within a proprietary software environment can lock an organization into that environment. Rules created in one environment do not transfer easily to another environment.
• KBESs can inhibit progressive thinking. Once a problem solving methodology or a design process has been implemented and personnel rely on it to automate tasks and quickly train new employees, organizational thinking becomes locked into the process.

Such organizational hesitance belies the ongoing boom in informal KM through software tools such as Lotus Notes and office automation software. For example, corporations employ spreadsheets to model, capture, and automate relationships and processes. They employ word processors to capture and communicate the justification for the processes selected. Under closer scrutiny, therefore, KM already represents a huge, dynamic market.

Unfortunately, corporations engaged in informal KM usually implicitly document product knowledge. They typically fail to make this knowledge explicit, or to organize it in a manner that can be leveraged for subsequent product iterations. When corporate culture instead provides the correct, explicit environment, KBESs provide the opportunity to effectively leverage information.

Currently, KBESs prove most effective as Rules-Based Systems that automate narrow, but well-characterized tasks. These tasks include the activities and methodologies of design, engineering, and manufacturing planning processes. The scope of rules can range from low-level modeling procedures employed with CAD systems, to selection of manufacturing processes based on component or assembly features.

Level 1 Knowledge Capture: Automation Of Narrow Tasks Within Expert Domains

When organized into effective building blocks, well-defined, narrow tasks can combine into increasingly complex processes that have a greater impact on the productivity of an organization. Examples of tasks relying on Level 1 knowledge capture include geometric modeling, finite element modeling, and machine tool path generation.

Level 2 Knowledge Capture: Automation Of Model And Data Abstraction

Slightly more advanced, Level 2 refers to the abstraction of models from one application to support others. Examples include idealization of design geometry for finite element modeling, or modifying the geometry to support more effective machine tool path generation. CAD/CAM/CAE suppliers support Level 2 by enabling the automatic conversion of design geometry into representations for other value-add applications. They support abstraction capabilities through a variety of tools including feature suppression, midsurface extraction, programming tools, and Application Programming Interfaces (APIs).

Level 3 Knowledge Capture: Automation Of Well Documented Design Processes

Level 3 includes the ability to capture and automate well-understood, often-repeated design and manufacturing processes. Effective Level 3 capabilities can employ the Level 1 and Level 2 automation as building blocks. For example, the design of an automobile radiator hose might involve a set of narrow, but well-characterized Level 1 and Level 2 geometric modeling and engineering analysis tasks. Level 3 combines these Level 1 and Level 2 "chunks of automation" into a process. The user defines critical parameters such as the hose radius and end locations. The software then guides the user through each decision, and automatically performs the required geometry creation and engineering analysis steps.

Knowledge Technologies International (KTI) and Prescient Technologies are the two most prominent, general-purpose, Level 3 KBES providers that support process modeling and automation. KTI develops and distributes its Knowledge Server and the Knowledge-Based Process Modeler. Knowledge Server—formerly referred to as the ICAD system—is a design process automation tool, integrated with leading CAD systems. KTI describes The Knowledge-based Process Modeler (KPM) as a "central nervous system," which integrates process modeling, workflow, and project management for their customers knowledge network. Prescient Technologies offers the STONErule application development toolkit to build customized KBESs. The STONErule environment includes object-oriented tools and database links that enable users to include material properties, manufacturing requirements, standards and regulations needed to facilitate the creation of rules-based design systems.

Many other software providers support Level 3 knowledge capture. Although they focus on delivering expert domain applications, they also recognize the value of such process-centric capabilities. For example, Dassault Systems, Parametric Technology Corporation, Structural Dynamics Research Corporation (SDRC), and Unigraphics Solutions provide a variety of process-centric Level 3 applications. Most recently, Unigraphics Solutions introduced MoldWizard, an application that guides users though the process of designing tooling for the manufacture of mold-injected plastic parts.

Level 4 Knowledge Capture: Automation Of Finding Solutions To Design Challenges

Over the past few years, Level 4 has gained increasing recognition. While Level 3 automates well understood design solutions, Level 4 automates the process of finding those solutions. Invention Machine's flagship Level 4 product, TechOptimizer, helps engineers resolve design problems through the support of an extensive library of proven scientific principles, including numerous application examples. The user describes a unique design problem, and the software employs semantics processing to analyze the problem description and attempts to retrieve the scientific principles that best apply as a solution.

Improving KM Capabilities

Given the broad variety of KM capabilities already available, whether through CAD/CAM/CAE systems or through office automation software, many corporations do not have to make additional software investments to improve their use of product knowledge. Instead, they can improve their effectiveness by applying guidelines like these:

• Establish a culture of knowledge management;
• Implement evolutionary, not revolutionary KBES deployment;
• Create a cross functional KBES team;
• Provide systems-level and component-level views, both of which knowledge capture requires;
• Employ the time of domain experts sparingly; and
• Use proactive promotion for KBES.