Validating a Commercial Toolkit to Evaluate and Implement Lean Manufacturing in High Mix Low Volume Facilities

Date

2022-05-12

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

For manufacturing facilities operating in a high mix low volume (HMLV) environment, applying the traditional methods of lean and the Toyota Production System (TPS) typically do not yield desired improvement results [1], [2]. Facilities in HMLV environments are required to be both responsive to external dynamic market conditions and flexible to quickly adapt to changing customer demand. Enabling a facility to be both responsive and flexible involves the following six dimensions: Process, Setup, Routing, Mix, Volume, Facility [2]. Prior research has examined specific dimensions in-depth, and many heuristic algorithms have been developed around the three NP-hard problems inherent to HMLV environments: clustering, layout optimization and scheduling. Many researchers have examined controlled cases studies on a few parts and machines. However, their narrow focus provides an incomplete view of the true complexities of a facility and may lead to sub-optimal solutions when put into context with the entire system. A holistic approach, for example Job Shop Lean (JSL), is required to examine each dimension, all part families and all routings to reach a feasible solution for the entire system given the constraints [2]. Our research provides a complete validation of JSL by executing the method under real world constraints at a large HMLV facility investigating over 5000 parts, 38 machines and 22 part families. We provide a comprehensive software toolkit for implementing JSL from initial data analysis, part family and machine clustering, layout analysis, scheduling, and modeling and simulation. The software toolkit we developed and models we integrated can be leveraged by practitioners to apply directly at their facility. In our case study, we reduced travel distance by about 50%, reduced wait time by about 150 hours and improved OEE performance by approximately 14% compared to the baseline in the original dataset.

Description

Keywords

Job Shop Lean, JSL, High Mix Low Volume, HMLV, Integrated Simulation Optimization, ISO, Python, AnyLogic

Citation