The objective of this award will be to conduct a one day workshop with the aim of clearly defining a research program for capturing, representing, and modeling mechanical engineering design knowledge that the research, academic, and industrial community can pursue in both the near and long term. A formalized language of mechanical engineering is envisioned that will allow engineers to communicate more precisely between each other and with computers. The resulting formalized ME design knowledge and language will be encoded and implemented in an open-knowledge repository. Such a knowledge repository could capture the disparate and representationally diverse ME knowledge of typical undergraduate students providing numerous benefits, including: 1) a standard for ME knowledge; 2) a knowledge base to support engineering design; 3) a knowledge base for computer-aided tutoring systems; and 4) a key component of the ME research and education infrastructure. A convergence of artificial intelligence, engineering informatics, description logics, and the semantic web with mechanical engineering design research will be the enabling factor to realize the vision of this research proposal and workshop.
If successful, the work may impact US economic and national security since a ME knowledge repository could increase the value of US graduates who have expertise beyond just the application of engineering science principles. Successful definition of a research agenda toward an ME knowledge repository has the potential for transformational outcomes. Increased awareness in the research community of high-impact research areas will increase interest in knowledge modeling. Establishing an ME knowledge repository as part of the educational infrastructure enables myriad methods and tools. The knowledge repository also has the potential to be a valuable resource for ME education research, as well as ME research in general. The networking function of the workshop will establish new relationships between individuals spanning widely across the field. This opportunity will provide a stimulus for new interactions and collaboration to advance the state of the art in engineering knowledge modeling. An understanding should emerge of the potential for this work to impact critical areas of engineering education, engineering information modeling, knowledge modeling, and the semantic web. The results will be disseminated through presentations, conference papers, and journal articles.
The primary objective of this workshop was to clearly define the needs and develop an outline for a research program for capturing, representing, and modeling mechanical engineering (ME) design knowledge. The idea is to model knowledge about ME principles, for instance the conservation of mass and energy, and their application to ME devices and systems using standard techniques such as free-body diagrams or control volumes. Further, the knowledge about how to formulate models of devices and of design decisions should be modeled. A major outcome of the study is a genesis of a research agenda that can lead to a ME Knowledge Repository encapsulating the knowledge of typical undergraduate ME students. Capabilities and limitations were identified of ontologies and ontology development methods and tools. Research approaches to overcome these limitations were suggested and focused on modeling knowledge about standard mechanical devices as well as knowledge about how to design such devices. Key knowledge types that are not clearly representable with these and other traditional methods should be further identified. Specific research agendas were identified in the five main areas of: design processes, the semantic web, knowledge representations, formal methods for representing knowledge, and modeling the meaning of mathematics. The technical exchanges among attendees fostered technical collaboration with the goal of enhancing research and development through cross-fertilization of technical interest areas. A research program was proposed that embodies the key elements of the research agenda. The program is focused around formal representations for languages for engineering functions, behaviors, and geometry, as well as the relationships among these languages. If the research program was conducted successfully, researchers could develop a ME knowledge repository that possessed the knowledge of a graduate of a mechanical engineering program. Such a knowledge repository would capture the representationally diverse ME knowledge domains by providing several benefits: 1) a standard for ME knowledge, 2) a knowledge base to support engineering design, 3) a knowledge base for computer-aided tutoring systems, and 4) more generally a key component of the ME research and education infrastructure.
