Investigates the collision-free path planning problem for a robot with two arms cooperating in 3-D work space. We have shown that motion constraints placed on the given robot can be equivalently represented as motion constraints placed on the end effector of the robot if two robot arms simultaneously hold the same object. By using this representation the given robot is shrunk to a point with its position being represented by its end effector and the obstacles are represented as forbidden regions in the work space. Thus, the problem of planning the motion for a robot is reduced to that of planning the motion for its end effector. The mapping function which maps the obstacles into their corresponding forbidden regions has been derived. We have proved that the forbidden region of an obstacle can be computed by mapping the boundary of the obstacle with the use of the derived mapping function. A potential field which is governed by a harmonic function in the 3-D work space has been computed and its gradient has been utilized in searching for the collision-free path. Simulation results are reported to verify the proposed method.

The authors present a new way to organize the codebook for vector quantization of speech. The new structure can reduce the storage requirement almost by half without any increase in the computational cost. The key is to store only the final code vectors in a tree structure. To facilitate the search at each node, one stores the indices that point to the code vectors that lie nearest to the other branches instead of the centroids of branches. The performance of the new method is compared with that of the conventional method. The new method offers a better compromise for applications that have limited computational resources and storage room and have less stringent distortion rate requirements.>

A description is given of the Robot Controller Test Station (RCTS), a software environment for implementing, testing, and evaluating robot control and sensor algorithms in both a simulated and a real robot system. RCTS is designed for flexibility, portability, ease of modification and ease of use. The authors discuss many of the design and implementation issues for RCTS. They outline the steps that a researcher can follow in order to integrate a control, sensing hardware interface or data analysis algorithm into an RCTS-based robot system. These steps do not require the researcher to be an expert programmer or even to comprehend the remainder of the robot software, so he or she can concentrate on algorithm design and optimization.>