Modeling framework to support decision making and control of manufacturing systems considering the relationship between productivity, reliability, quality, and energy consumption. (January 2022)

• Record Type: Journal Article
• Authors: Saez, Miguel; Barton, Kira; Maturana, Francisco; Tilbury, Dawn M.
• Is Part Of: Journal of manufacturing systems. Volume 62(2022)
• Page Start: 925
• Page End: 938
• Publication Date: 2022-01
• Subjects: Smart manufacturing -- System-level control -- Multi-objective optimization; Manufacturing processes -- Periodicals; Production engineering -- Data processing -- Periodicals; Robots, Industrial -- Periodicals; Production, Technique de la -- Informatique -- Périodiques; Robots industriels -- Périodiques; Electronic journals; 670.42
• DOI: 10.1016/j.jmsy.2021.03.011
• Languages: English
• ISSNs: 0278-6125

Abstract

Highlights: A framework for modeling and simulating manufacturing systems that captures the relationship between productivity, reliability, quality, and energy consumption. The framework is based on a hybrid model of machine-level dynamics combining discrete states and continuous variables with system-level manufacturing performance. The formulation of a multi-objective optimization problem to support opportunistic decision making. The problem is formulated to evaluate both system-level variables such as production line configuration and machine-level variables such as speed or feedrate. The modeling and optimization framework is demonstrated in a case study based on a fully automated manufacturing testbed. The framework is used to define control actions for machine-level operation and system-level configuration in a flow shop system. Abstract: Different formalisms and modeling frameworks have been developed to capture the discrete behavior of manufacturing systems. However a purely discrete model does not capture the continuous variables of machine-level operation. Previous research in modeling manufacturing systems has studied the quantity-quality and quantity-reliability coupling of production environments but their relationship to energy consumption is often not considered. This work presents a framework that extends the state-of-the-art in modeling manufacturing systems to support decision making for control at both machine- and system-level variables using hybrid models. The modeling strategy considers the coupling between productivity, quality, reliability and energy consumption and leverages the current plant floor data extraction capabilities to develop data-driven models. A novel control strategy is presented that considers the effect of various machine- and system-level variables over different performance metrics and balances them using multi-objective optimization. This framework was validated using a combination of real and simulated data of a production representative environment. The optimal set of control variables is obtained using simulation-based optimization to support plant floor decision making by studying process variables, maintenance actions, or system reconfiguration. Results show the effect of different control variables and the ability to reduce energy consumption while improving productivity. The implementation of the modeling and control framework presented here has the potential to impact the operations of manufacturing system by reducing cost and improving productivity.