Geometric Constraint Solving
Product design for manufacture is an activity driven by descriptive information. Increasingly, design includes the use of design constraints that impose conditions on the shape of the product. That is, the designer states specific constraints without telling the system in detail how to satisfy them. It is the task of the underlying constraint solver to derive a plan by which to solve the constraint system.
Constraint solving will be presented in two parts. In the first part, the architecture of a simple planar constraint solver is described that anyone can implement. We then discuss ways in which to extend the solver, both regards geometric coverage as well as more complex constraints patterns.
The second part addresses spatial constraint solving focusing in particular on the algebraic side and some of the techniques that are known for solving the equations.
Before joining the Purdue faculty, Professor Hoffmann taught at the University of Waterloo, Canada. He has also been visiting professor at the Christian-Albrechts University in Kiel, West Germany (1980), and at Cornell University (1984-1986). His research focuses on geometric and solid modeling, its applications to manufacturing and science, and the simulation of physical systems. The research includes, in particular, research on geometric constraint solving and the semantics of generative, feature-based design. Professor Hoffmann is the author of Group-Theoretic Algorithms and Graph Isomorphism, Lecture Notes in Computer Science, 136, Springer-Verlag and of Geometric and Solid Modeling: An Introduction, published by Morgan Kaufmann, Inc. Professor Hoffmann has recieved national media attention for his work simulating the 9/11 Pentagon attack.
He is on the editorial boards of