Web-based Group Decision Support System for Solving Assembly Line Balancing Problems

Abstract: In the automotive industry, assembly lines are used to produce vehicles. These assembly lines improve throughput, and need to be carefully planned. Planning, or balancing, an assembly line constitutes identifying precedence relationships between tasks in the assembly line, and assigning tasks to stations to fit some criteria. This procedure is costly to do by hand, and is well-suited for some level of automation. The problem of balancing assembly lines has been researched since the 1950’s, but modern assembly lines largely rely on engineers to balance the line by hand. This thesis proposes that the work flow of engineers planning the assembly line would be improved by a group decision support system. This group decision support system could supply engineers with proposals for assembly lines, which the engineers can choose to modify further, either by hand or with the decision support system. The group decision support system is realized with a distributed system, consisting of a front-end, a back-end, an application programming interface to balance assembly lines, and two databases. The front-end is a website, where the users can create problems with a precedence graph. The back-end allows data to be permanently stored in the two databases, and allows communication with the application programming interface. The contribution of this thesis is a proof of concept of a group decision support system that can solve two basic types of assembly line balancing problems, SALBP-1 and SALBP-2. During development it was found that the developed system was generic enough to support different types of tasks, such as planning cooking. Further development is needed to use the system in an industrial setting, as real assembly lines need more complex models than the current version support.