In robotics we are interested in motion planning with emphasis on high-dimensional systems and kinodynamic planning, mobile manipulation, multi-robot systems, hybrid systems, assembly planning, reasoning with sensing and control uncertainty, physical modeling, flexible object manipulation, probabilistic methods in robotics, the geometry and physics of motion, and the use of new enabling technologies.
In bioinformatics and biomedicine we develop computational tools on high-performance systems to model protein structure and function, understand bimolecular interactions, develop new medicinal drugs, and help analyze, in the long run, the molecular machinery of the cell.
Both areas above involve real-world problems and fall into the broader category of physical computing. In both areas we seek to develop physical algorithms: algorithms that are capable of solving complex high-dimensional geometric problems arising in real-world applications (e.g., move a robot from A to B, predict a biomolecular complex). We believe that as computers become ubiquitous, we need to use computers to represent, simulate, and interact with the physical world. This is not an easy task, however. Algorithms for physical problems differ in significant ways from those for traditional (artificial world) problems. The latter algorithms have full control over and perfect access to the required data. In contrast, physical algorithms apply to objects in the real world which are subject to the independent and imperfectly modeled laws of nature. Our long term goal is to study the fundamental issues arising when algorithms are designed for problems in the physical world and to develop coherent solution frameworks which quantify, to the extent possible, the tradeoff between accuracy and performance present in solutions developed for realistic settings.
For more information read: Short Description of Research Activities
Our research is currently supported by NSF, NIH, the John and Ann Doerr Fund for Computational Biomedicine, Shell, Willow Garage, and a Sloan Fellowship. For our work, we use high-performance computer clusters supported by NSF in partnership with Rice University, AMD and Cray. Part of our work is supported by the National Science Foundation through TeraGrid resources.