Toward the long term goal of developing computational tools for dealing efficiently with the complexity and the constraints of physical tasks, the research targets the solution of several well defined robotics problems and the improved understanding of algorithmic techniques that are useful for applications arising in the physical world. The core areas of research include the application of randomized methods to geometric problems, the incorporation of physics-based models and techniques in geometric frameworks, and the development of data structures for storing/searching geometric data and for supporting continuous motion. A few key problems have been identified to guide and inspire research in the above domains. These are: assembly with emphasis on virtual prototyping for maintainability studies involving rigid and flexible parts, path planning in cluttered environments, and manipulation with emphasis on part orientation. Practical solutions will be sought for the above problems. The proposed research plan thoroughly integrates educational activities at Rice University. The main focus of these activities will be (a) the immersion of students, and especially undergraduates, into research projects, (b) the introduction of a course in robotics and the restructuring of basic theory courses to include algorithmic techniques that have been proven useful in graphics and robotics, (c) the mentoring of women students, and (d) the promotion of industrial and interdisciplinary training for Ph.D. students.
