TRACS: The Hardware and Software Architecture of a New Two Robotic Arm Coordination System

This paper presents the hardware and software architecture implemented in the Two Robotic Arm Coordination System (TRACS) at the GRASP Lab of the University of Pennsylvania. It is developed to perform experiments on dynamically coordinated control of multiple robotic manipulators. Its architecture avoids complexities and allows the user to easily implement desired control algorithms. This system controls two PUMA 250 robot manipulators, each with 6 DOF. The IBM PC-AT is chosen as the host computer because of its ease in real-time programming, simplicity of I/O interfacing, and low cost of hardware and maintenance. The Intel 286 processor of the PC-AT is aided by a AMD 29000 high speed floating point processor based board. Together, the 286 provides the real-time environment and performs sensor and manipulator I/O while the AMD 29000 calculates the real-time control algorithms. TRACS incorporates MO, a C library of routines being developed in the Grasp Lab to control robots. MO separates hardware dependent software from hardware independent code and provides the user with a virtual robot interface. End-effectors are built to perform two arm grasping and manipulating of large objects. The end-effectors are outfitted with contact/force sensors. The system is capable of controlling of the two cooperative manipulators at 200 Hz. Comments University of Pennsylvania Department of Computer and Information Science Technical Report No. MSCIS-90-70. This technical report is available at ScholarlyCommons: http://repository.upenn.edu/cis_reports/559 TRACS: The Hardware And Software Architecture Of A New Two Robotic Arm Coordination System MS-CIS-90-70 GRASP LAB 236 Eric Paljug Xiaoping Yun Filip Fuma Department of Computer and Information Science School of Engineering and Applied Science University of Pennsylvania Philadelphia, PA 19104-6389