Abstract

Two important trends in developing innovative and efficient approaches for improving plant productivity are cellular manufacturing and robotics. The proliferation of robot technology is an outcome of increasing industrial automation especially in engineering and electronics. Robots offer substantial gains in manufacturing productivity, particularly when integrated into an automated system. Robotic cells involve the use of robots to feed machines in manufacturing cells. The factors affecting the performance of such systems include sequencing robot moves, sequencing parts, buffering, and cell design. This thesis addresses the problem of sequencing robot moves in a two machine manufacturing cell in the presence of a buffer. We develop cycle time formulae using a state space approach. We adopt analytical methods for determining the optimal cycle time of a two-machine robotic cell with a single buffer producing identical parts. We also evaluate the effectiveness of buffering in reducing the cycle time. We extend our research to the robotic cell producing multiple part types. We consider the production of a quantity known as minimal part set (MPS) for multiple part types, to be compatible with the recent trend toward just-in-time manufacturing. Our objective is to identify optimal robot move sequences for a pre-determined arrangement of parts in a minimal part set. We accomplish our goal by developing a branch-and-bound algorithm. We also provide a comparative analysis for scenarios with and without a buffer to establish the usefulness of a buffer.