Application Of APS In Planning And Scheduling In Automobile Production Line

Since the late 1990s, the auto industry is trapped in the vortex of ever-expanding reduction of inventory, maintaining market share. War of attrition between the major automobile manufactures results in continually deterioration of profits of entire auto industry, including auto factories and accessory suppliers.Advanced Planning and Scheduling (APS) is a constraint-based scheduling with limited production capacity. It is an optimization technology under resource constraints, used for both internal planning and scheduling within single enterprises in short-term and long-term forecast under known conditions between enterprises. Its unique advantages make it attract more and more attentions from automobile manufacturers. These manufacturers hope to use the APS to improve their production, and raising the efficiency of enterprises.Theory of Constraints (TOC) is rooted in the Optimized Production Technology (OPT), which focuses on the improvement and a set of concepts and management principles about how to best implement these improvement. OPT acclaims that the goal for a business's plan and control is to seek the best coordination of demands from clients and enterprises'ability. Once a controlled process (that is the bottleneck) has established a dynamic balance, the remaining procedures should be synchronized with the corresponding process. OPT's plans and control is achieved through the DBR system. Bottleneck resources use two-side scheduling, while non-bottleneck resources use inverted planning. We can attach buffer between any complex resources. The DBR logic is used to synchronize all the resources and materials of key processes.Car production can be divided into four parts, that is stamping production line, soldering production line, varnishing production line and assembly production line in accordance with the sequence of the production process. The most important concept in production line is the meter, which stands for a fixed length of time. Production is undergoing in the unit of meter, meter controls the tempo of the production process. We can see from this that the concept of bottleneck in TOC has been balanced with meter. Through daily production plan, so called the plan after the assembly production line, we can get the plan before the assembly production line, so called the plan after varnishing production line; furthermore, we can get the plan before the varnishing production line, and so on and so forth. In accordance with the idea that reversing push can we come to the production process plans.The first difficult problem to solve during the scheduling process is about the circuitous process in the varnishing production line. Circuitous process will be divided into three parts: the first painting time, which is referred to the first processing time in the workflow and the second painting time, which is referred to the second processing time in the workflow.Circuitous preparation time, which is referred to the preparation time between the two painting process. The basic objective of sorting is to make the varnishing production line most effective and a circuitous preparation process, in order to maximize efficiency of the use of varnishing production line, and provide the smoothest possible production process.Sorting algorithm of the circuitous process is based on the idea of reversing push: get the plan before the varnishing production line out of the plan after the varnishing production line. Check from the tail of the varnishing production line, one car on each check. If the car is needed for a secondary varnish and has finished painting, mark it and moving it forward to a position of the former location where it should be before varnishing. The distance of forward movement of the car equals to the time span needed for the first painting time. At the same time, the time used for the first varnish of the car should be marked as used, preventing other cars from using this time.The second difficult problem to solve during the scheduling process is about the converting from the batch process into the flow process. We divide the conversion into two parts: calculation of the time of components in the workflow as well as calculation of the work center. For calculation of the time of components in the workflow, further calculation can be based on the sequence before the varnishing production line, and then by examining properties of each work center in the soldering assembly line can come the result of the demand of the varnishing production line.We need to consider two types of the work centers: one type of work center needs only to produce one single type of component, while the other type of work center needs to produce two different types of components. The first type of work center could even theoretically be omitted its buffer between corresponding work space in the soldering production line and the work center. The second type of work center needs preparing-time in order to change processing tools used for processing the other type of component. To minimizing the time used for changing tools, there is a need for batch processing, the same parts produced in a batch. We should make full use of buffer, so that batch quantity is to be the greatest. For transit process, the purpose of reducing mass transit is to improve the continuity of the production process, and to balance the overall process. Small transit quantity results in rise of cost of transition. Assuming a work center WC, its daily production of components needs to meet N cars, its batch transit quantity is qt. Buffer size is k times of its batch transit quantity, and buffer safeguarding each set of the component of the car costs cb. Cost in each transfer process is ct, transit daily total cost is C. To minimize the cost, we need to calculate the minimum value of C=ct×?? N/qt?? +cb×qt×k as well as satisfy the relation of qt×k<=QB. For the process of seeking the minimum value, we use the method of test-optimization, which is constantly on the test of values when probe to find the smallest value of qt and k.Finally we implement an APS system simulating the auto production line scheduling, combining B/S and C/S mode and share a same database. First, use B/S mode to simulate several dealers/customers. Then use random procedures to make orders and store them into the database. APS system preliminarily sort confirmed orders of the day in accordance with the different areas of the order to facilitate the transportation of after production. The sequence is used as the sequence of the plan after assemble production line as well as the plan before assemble production line. Subsequently we use reverse sorting algorithm to calculate the plan before the varnishing production line, which is also used as the plan after soldering production line as well as before soldering production line. Other working centers use this as the basis for conduct scheduling. Then we use test-optimization algorithm in calculating demanding quantities and time for components. Stamping workshop schedule is largely the same as schedule made in varnishing workshop.Experimental results show that, ideas and algorithms promoted in this paper are practical and of high efficiency. For future work, on one hand, we need to promote the APS system designed in this paper into practical scope and have the system constantly revised and improved; on the other hand, we need to modularize and componentized the parts of the system so that the system based on this paper can be appropriate in more types of production line and in more types of production process, bringing more flexibility, more extensive and more versatile.