Markovian Reliability in Multiple AGV System

Background

Flexible manufacturing systems (FMS) are crucial for modern manufacturing to enhance productivity involved with high product proliferation. As one of the critical components of the FMS, the flexible material handling system (MHS) plays a strategic role in the implementation of the FMS. According to Tompkins et al. (2002), about 20–50% of the total production cost is spent on material handling. This makes the subject of material handling increasingly important. In addition, all the complexity of manufacturing are passed on to the MHS. Therefore, the flexible MHS has been vital for improving the FMS to fulfill the requirements of high product proliferation (Zhao et al., 2011).

Automated manufacturing systems (AMS), which are equipped with several CNC machines and AGV-based material handling system are designed and implemented to gain the automation and efficiency of production. To make use of all features of AMS, the planning in the AMS decision making process is critical because the planning decision has influence on the subsequent decision processes such as scheduling, dispatching, etc. The planning in automated manufacturing systems can be characterized as being online and short-term nature to respond to frequently changing production order. Given a production order, manufacturing planning function is responsible to establish a plan by decomposing the production task into a set of subtasks. An analysis of AMS dealing with changing demand can be found in (Terkaj et al., 2009). An extensive review of the loading problem for an FMS can be found in (Grieco et al., 2001). An early stochastic programming approach to address the short-term production planning for an FMS can be found in (Terkaj & Tolio, 2006).