Toward data-driven production simulation modeling: dispatching rule identification by machine learning techniques

Published

June 24, 2019

Author(s)

Satoshi Nagahara, Timothy A. Sprock, Moneer Helu

Abstract

Production simulation is useful to predict and optimize future production. However, it requires much effort and expertize to create accurate simulation models. For instance, operational control rules such as job sequencing rules and resource assignment rules are modeled based on interviews with shop-floor managers and some assumptions since those rules are tacit in general. Since operational control rules determine dynamic behavior of production systems, it is important to model the rules accurately. In this paper, we consider a data-driven approach for modeling operational control rules. We develop job sequencing rule identification methods that models the rule automatically from historical production data by machine learning techniques. The methods are evaluated through computational experiments using actual factory data in terms of accuracy and robustness against uncertainty in human decision making.

Proceedings Title

Procedia CIRP

Volume

81

Conference Dates

June 12-14, 2019

Conference Location

Ljubljana, SI

Conference Title

52nd CIRP Conference on Manufacturing Systems

Pub Type

Conferences

Download Paper

https://doi.org/10.1016/j.procir.2019.03.039

Local Download

Citation

Nagahara, S. , Sprock, T. and Helu, M. (2019), Toward data-driven production simulation modeling: dispatching rule identification by machine learning techniques, Procedia CIRP, Ljubljana, SI, [online], https://doi.org/10.1016/j.procir.2019.03.039, https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=927446 (Accessed May 5, 2024)