Our goal is to develop a science base for the modeling and simulation of real world systems. To this end, we are pursuing research in object representation, solid modeling, graphics, and automated design. Some of our most significant research has been in the area of robust geometric algorithms. We have developed Newton, a general-purpose, model-driven simulation system that is object-oriented. It is based on a modular design, thus allowing the incorporation of different modelers as needs change. As an outgrowth of Newton, we are now developing an environment and toolbox, SimLab, that will allow users to build their own simulators. Other areas under investigation are guaranteed-quality finite element meshes, automatic generation of software that numerically integrates systems of ordinary differential equations, and gripping and manipulation. In collaboration with the Design Research Institute, we are developing technologies to facilitate information capture and access within an engineering design environment. Among the technologies being researched are distributed databases and persistent object storage, document image processing and management, multimedia and user interface technology, information science for heterogeneous data, knowledge representation and organization, and remote collaboration technologies.