A Study on Bottleneck Identification Method for Fully Manual Mixed-Model Assembly Lines

DOI: https://doi.org/10.62517/jiem.202603303

Author(s)

Hongbo Man, Wei Wang*

Affiliation(s)

College of Mechanical and Electrical Engineering, Northeast Forestry University, Harbin, Heilongjiang, China *Corresponding Author

Abstract

Fully manual mixed-model assembly lines suffer from significant skill variation among operators and dynamic bottleneck drift caused by changing product mix ratios. Traditional static time study methods deliver low identification accuracy in such production environments. To address these challenges, this paper proposes a four-step bottleneck identification method: static preliminary screening, skill correction, dynamic verification, and drift analysis. The proposed method eliminates efficiency disturbances caused by individual operators by constructing a product-specific skill coefficient matrix, establishes a macro-micro dual verification system to identify genuine bottlenecks, and calculates the critical conditions for bottleneck drift via mix ratio sensitivity analysis. We validated the proposed method using real-world tests on a refrigerator assembly line and multi-scenario FlexSim simulations. The method achieves an overall identification accuracy of 93.8%, identifies the critical product mix ratio for bottleneck drift as 77.8%, and cuts the misjudgment rate of pseudo-bottlenecks to less than 5%. Requiring no digital hardware infrastructure, the method cuts the implementation cycle by approximately 70% compared with full-process simulation, and provides a lightweight optimization solution for mixed-model production capacity improvement applicable to small and medium-sized manufacturing enterprises.

Keywords

Fully Manual Mixed-Model Line; Bottleneck Identification; Skill Coefficient; Work Measurement; Bottleneck Drift

References

[1] Kumar S, Singh R. A Review of Static Time-Based Methods for Bottleneck Detection in Assembly Lines. International Journal of Production Research, 2024, 62(4): 1120-1138. [2] Zhang Y, Liu M, Xu L. FlexSim-Based Simulation for Bottleneck Validation in Manual Assembly Lines. Simulation Modelling Practice and Theory, 2023, 125: 102731. [3] Liu Q, Wang F. A Light-Weight Bottleneck Identification Method for SMEs without MES: A Field Study. IEEE Transactions on Engineering Management, 2024, 71: 3456-3468. [4] Li J, Zhao X. Real-Time Dynamic Bottleneck Identification without IoT Infrastructure: A Motion-Study-Based Approach. Journal of Intelligent Manufacturing, 2024, 35(3): 1299-1315. [5] Lu Y, Zhang X, Wang J. Dynamic Bottleneck Identification in Mixed-Model Manual Assembly Lines Using a Skill-Adjusted Workload Model. Journal of Manufacturing Systems, 2023, 68: 234-246. [6] Sun H, He Y. Pseudo-Bottleneck Discrimination in Manual Mixed-Model Lines Using Time Study and Skill Weighting. International Journal of Industrial Ergonomics, 2025, 95: 103489. [7] Park J, Kim D. Bottleneck Drift Analysis in Mixed-Model Production: A Critical Ratio Approach. Production Planning & Control, 2025, 36(1): 45-59. [8] Lee J, Park H. Critical Mix Ratio for Bottleneck Transition in Two-Product Manual Assembly Lines. Journal of Manufacturing Processes, 2024, 99: 1-12. [9] Yang J, Chen Y. A Hybrid Method Using Standard Time and On-Site Observation for Bottleneck Identification in Low-Digitalization Factories. Robotics and Computer-Integrated Manufacturing, 2025, 87: 102697. [10] Chen W, Li H, Zhao Q. Operator Heterogeneity Modeling Using Skill Coefficients in Manual Assembly Systems. Computers & Industrial Engineering, 2023, 176: 108945. [11] Wang Z, Zhou T. In-Process WIP Distribution as a Macro Indicator for Bottleneck Confirmation: A Case Study. Assembly Automation, 2023, 43(2): 210-222. [12] Kim B, Lee S. Operator Learning Curve Effects on Bottleneck Shift in Manual Assembly: A Simulation Study. International Journal of Production Economics, 2023, 260: 10883