A loop material flow system design for manufacturing plants

grantor: University of TorontoAround 20 to 50 percent of the operating expenses within manufacturing plants are related to material handling. Effective facilities planning can reduce these costs by 10 to 30 percent. The material flow system is the foundation of facilities planning. It is defined in terms of the type of the material handling equipment, the flow network, its directions, and the number and locations of Pick-up and Delivery stations. Automated Guided Vehicles are now considered as the basic material handling equipment in manufacturing plants. They are preferred to conveyors because of their flexibility, and to robots because of their mobility. We develop a model to design a unidirectional loop flow system for automated guided vehicles. A set of assumptions are examined. Linear Programming, Integer Programming, and Network Flows are modeling tools of our research. The shortest loop problem is transferred into the Generalized Traveling Salesman Problem. Compared to available models for designing a loop on a block layout, our model is better integrated, has a smaller number of binary variables and stronger constraints. A Goal Programming approach to design the shortest loop with a secondary objective function is also examined. The shortest loop model minimizes all costs proportional to the length of the loop, but does not cover the travel costs. We integrate both the design of the unidirectional loop, and determination of the number and location of stations in order to minimize the total loaded vehicle travel distance. No earlier model designs the loop and the station locations simultaneously. To solve the problem, we have concentrated on 3 directions. Having a stronger LP-relaxation, an intelligent Branch and Bound, and a set of MIP solver methods which are most appropriate for this particular problem. The model is extended to minimize the cost of the total system, including the network, stations, and loaded travel distance. Capabilities of the model to cover the conventional configuration and the bi-directional loop are also discussed. Finally the model evaluates the impact of the empty vehicle travel. All considerations of the loaded travel distance formulation and solution procedure, as well as a decomposition approach are implemented to solve the model. We show that the optimal solution for loaded travel distance is far from optimal for both loaded and empty travels. The models developed in this research may be examined in three environments: fixed loop-variable station location, variable loop-fixed station location, and variable loop-variable station location. However, it is mostly implemented in the last environment. Computational results are reported for a set of randomly generated problems as well as some well known examples.Ph.D