Modeling and Analysis of Global Vision for Operating Multiple Free-Ranging AGVs on the Factory Floor

点击翻译按钮获取中文摘要

A system to provide the information necessary for the control of multiple free-ranging automatic guided vehicles will be developed based on the use of stationary visual cameras, or what is termed "global vision," in order to expand the range of applications for which they are a high-performance and cost- effective transport alternative. The expected results of such a system include: increased transport throughput because vehicles can operate faster and globally efficient paths can be planned; improved safety due to increased collision detection and avoidance capabilities; and reduced vehicle costs. In order to demonstrate that global vision is technically feasible for vehicle control and is cost-effective as compared to other feasible means of control, the following tasks will be performed: (1) camera placement, where algorithms will be developed to determine a minimal number of cameras and their positions in a facility; (2) object tracking, where difference image-based methods will be developed to detect and track moving objects; (3) performance measurement, where actual hardware and software implementations will be used to demonstrate the technical feasibility of global vision for a typical factory shop floor; and (4) economic feasibility, where analysis techniques will be developed to compare different methods of controlling free-ranging transport vehicles.