Dynamic Production Line Re-balancing by Alternative Plans for Compensating Equipment Failures

As social uncertainty increases, there are changes in values such as safety, sustainability, and response to changes. The manufacturing industry is also required to improve its resilience by dynamically recombining production resources to respond to change. In the operation of a production line, if demand fluctuates due to changes in the external environment or customer needs, or if sudden production fluctuation such as equipment failure occurs, the production line will be stopped due to reconfiguration or equipment restoration. Therefore, the challenge is to continue production and improve productivity even when fluctuation occurs. In order to respond to the above challenges, this research aims to dynamically and quickly re-plan the production line using the existing equipment according to current equipment and production status. In this paper, failures of the production equipment, such as robots and tools, are regarded as production fluctuation factors. In order to continue production with the remaining resources in the event of equipment failure, it is necessary to change the process plan. However, since these are engineering tasks in the production preparation stage, there was the problem of not being able to flexibly change them at the manufacturing site during the production execution stage. In this development, an alternative plan pre-generation and selection method was developed to maintain production by re-allocating tasks during equipment failure. In this approach, first, multiple alternative plans for task allocation for failures in each piece of equipment are planned in advance using a process plan optimization technique. Next, production fluctuations due to equipment failures are detected. Finally, dynamic changes are possible by selecting alternative plans that maximize the throughput of the entire production system. Initial verification results are shown by comparison with the conventional method for equipment failure using simulation on a small-scale robot assembly line for automotive inverter.