9409716 Zhuang The award is for research to develop procedures for self- calibration of machine tool structures based on the Stewart Platform. The research would address issues of sensor location and desired accuracy, system kinematic modeling for purposes of calibration, and the design of optimal alogarithms for calibration and compensation. The self-calibration technique, if successful, would be of use to a new class of Computer Numerically Controlled machine tools based on the Stewart Platform. The accuracy of such machine tools would be improved as various types of machine errors, including those induced by thermal gradients and load variations, can be compensated for without interrupting normal machine operation.