Global vision for the control of free-ranging automatic guided vehicle systems

This dissertation introduces the method of "global vision" for the control of free-ranging automatic guided vehicle (AGV) systems. Global vision refers to the use of cameras placed at fixed locations in a workspace to extend the local sensing available on board each vehicle in a free-ranging AGV system. The fixed cameras provide the information to robustly control AGV systems composed of large numbers of low-cost vehicles. A system with these capabilities can expand the range of applications for which free-ranging AGVs are a high-performance and cost-effective material transport alternative. A possible overall global vision-based AGV control system is first described, and then detailed methodologies are presented for camera placement, mobile object tracking, and determining the economic feasibility of global vision-based vehicle control. The camera placement problem in global vision refers to the determination of a minimal number of cameras and their placement in the workspace. This problem is shown to be NP-hard through restriction to the two-dimensional NP-hard problem Minimum Star (or Guard) Cover of a Polygon with or without holes. A convex-hull-based greedy heuristic is described that provides reasonable camera placements. Global vision-based AGV control provides for increased collision detection and path planning capabilities due to the ability to track mobile objects located in regions beyond the range of an AGV's on-board sensors. The tracking performance of a difference-image-based Kalman filter is detailed, where the correspondence problem is greatly reduced through the use of reference images instead of successive images. The economic feasibility of global vision-based control, as compared to other methods of free-ranging AGV control, is estimated through the use of a model formulated as a transportation problem. Each method of control is distinguished by its required equipment costs and the level of collision detection capability it provides. The potential impact of global vision for AGV control includes increased system throughput, improved operational safety, and reduced equipment costs.