Balanceamento de linhas de montagem por meio de modelos de programação linear inteira mista: estudos em indústria automotiva

Abstract

In product layout manufacturing systems, or production lines, the elapsed time between the completion of two consecutive products affects the line productivity. This elapsed time is defined as the cycle time of the line and it is given by the workstation with greatest processing time within the line. In order to reach a productivity increase, the cycle time must be diminished and optimization methods that aim at balancing the production line are commonly used in this context. The line balancing acts to minimize the gap among the workstations processing times, aiming to reduce the processing time of the busiest station of the line, called the bottleneck station. This project takes into account a mixed-model assembly line balancing problem that features two special operational conditions: Automatic Tasks and Common Tasks. Automatic Tasks are the ones that are executed by automated equipment, without the need of action from a worker throughout its execution. Common Tasks are distinguished by requiring or benefiting from the simultaneous work of two or more operators. Both conditions can be frequently seen in automotive industries, as is the case of the assembly line that was focused on this paper. This assembly line was chosen because both Automatic and Common Tasks are featured on it. In order to get the desired results of the line balancing, a mathematical model in Mixed Integer Linear Programming (MILP), which indicates the optimum distribution of the tasks among stations and workers in the line, was developed. The goal in this work is to evaluate the influences that Automatic and Common Tasks have on the Assembly Line Balancing Problem and reach a feasible solution closer to the real operational conditions of the studied line. The results show that both Automatic and Common Tasks causes more changes to the workers time distribution than they do to the workstations time distribution. It was possible to see also that, within studied scenarios, the only way to achieve time cycle reduction with Automatic Tasks is to assign them to the bottleneck station of the line. In addition, it was observed that Common Tasks can increase the workers workload and raise the cycle time accordingly. Common tasks, though, can be imperative in a number of cases in industrial scenarios. At last, it can be said that the proposed model can solve non- trivial assembly line balancing problems as it takes into account the possibility of occurrence of both Automatic Tasks and Common Tasks in assembly lines.