The study of generative grammars as originally defined by Chomsky has extended beyond the original audience of computational linguists to a broader audience of researchers. They have been employed to model and automatically create music, engineering products, architectural buildings, and even Mondrian paintings. Prof. Matthew Campbell has concentrated his research on the automated synthesis potential of generative grammars in the past five years as a means to improve the efficiency and productivity of innovation in engineering design.
Towards this grand goal, several research issues are presented: first, the definition of a formalism capable of capturing the diversity and detail of engineering design knowledge; second, the manner in which the resulting space of potential solutions is searched; and third, the approach in deriving rules of the grammar and rapidly implementing and compiling them in the system. In order to explore these research topics, Matthew Campbell works at the Virtual Product Development Group in the Institute of Product Development at TUM. The goal of the collaboration is to define a fundamental computational theory to be used in developing a universal, open-source, shape and graph grammar synthesis system that provides a base platform for further research into the benefits and potential for automated design. The interdisciplinary research involves a mixture of engineering and design methodologies, computer science, artificial intelligence, and optimization fundamentals.
Prof. Peter Schröder is a Professor of Aeronautics and Mechanical Engineering at the California Institute of Technology (Caltech) and was as a Hans Fischer Senior Fellow active in this TUM-IAS Focus Group.
TUM-IAS funded doctoral candidates:
Jan Stühmer, Computer Vision and Pattern Recognition
Amir Hooshmand (PhD in 2014)