This was the focus of the first part of my research program which addressed the communication of CAD model knowledge in an engineering enterprise. This work has been implemented in part to support life-cycle facilities engineering at the DoD and presented at a conference in formal methods for CAD. The fundamental issue addressed by this research was that recent CAD frameworks for simultaneous engineering systems focused on the communication of CAD model information in the context of the interoperability of CAD tools rather than the communication of designers. I should note that the SHADE project by Tom Gruber, Mark Cutkosky and Michael Gensereth of Stanford University demonstrated how agents might facilitate communication between engineers using a common ontology and transport mechanism. However, they arrived at the knowledge sharing agreements before the design project even began, essentially a priori deciding what information they'll share. The SHADE project typifies current research in design which views design as a knowledge intensive process and stresses the importance of constructing knowledge representation formalisms to support design communication. But there is an orthogonal argument which came up during this year's ASME Design Theory and Methodology conference by Professors Ullman at Oregon State and Ward Seering of MIT and others who argue that while, perhaps design is a knowledge intensive process, the knowledge intensive portion lies not in product modeling and building knowledge sharing protocols but rather in understanding the design as it evolves to bring the right information with a high degree of confidence to the designer at the right time.
Go to the next slide.
Transition: The mechanical design language interpreter, or MDL for short, introduces a computable methodology to learn these associations based on the design text. My research program formalizes the learning process and considers metrics to measure how well the system learns design data.
| BEST Lab | Department of Mechanical Engineering | University of California at Berkeley