Evaluation of Production Systems Using Artificial Intelligence - Application to Doorlock Assembly System

• Main Authors: Nozaki, Yutaka, Ishikawa, Hiroshi, Asai, Susumu
• Format: Report
• Language: English
• ISSN: 0148-7191; 2688-3627
• DOI: 10.4271/950821

As a part of our process engineering (planning) steps and in order to shorten the initial flow control after the implementation of a production system, we have been adopting countermeasures for defects that have been forecasted by the application of FMEA. However, due to this having to be done with limited information (which is dependent upon the skill level of the operator) and due to the fact that the effects of each particular countermeasure are not fully understood, we cannot be certain that the measures implemented during the planning stages are thorough enough. The current situation is that countermeasures for most issues are being handled during the initial flow control stages based on the trial-and-error method. Given this situation, normally, it would take us more than three months from the time of line-off (start of production) to achieve our target rate of capacity utilization. Therefore, to keep the countermeasure period as short as possible, the number of malfunctions can be limited by forecasting the capacity utilization rate of the production system (equipment) after it has been implemented to a minimum level.