Processing time estimation errors' impact on dispatching performance

• 成果类型: Article
• 署名作者: Treleven, Mark D.; Watts, Charles A.; Penlesky, Richard J.
• 署名单位: University System of Ohio; John Carroll University
• 刊物名称: PRODUCTION AND OPERATIONS MANAGEMENT
• ISSN/ISSBN: 1059-1478
• 发表日期: 2006
• 页码: 144-155
• 关键词: job shop; dispatching; due dates; processing time estimation; simulation

摘要

This paper examines the impact of inaccuracies in the estimation of processing times on the performance of dispatching rules in dual-resource-constrained and machine-limited job shops. Four distributions of estimation errors are examined in eight different simulated shop environments representing DRC and ML job shops with different labor utilization levels and dispatching rules. A unique approach to setting total-work-content-based due dates is introduced and recommended as more appropriate than the approaches previously used. The mean and standard deviation of the absolute value of lateness are reported as performance measures. The results indicate that errors in the estimation of processing times have a significant impact on shop performance. Specifically, optimistic and high variance estimated processing time distributions produce significantly worse shop performance than accurate and pessimistic errors. The interactions between the effects of the estimated processing time distribution and due date tightness are also shown to be significant.