2015 Using heuristic algorithms to solve the scheduling problems with job-dependent and machine-dependent learning effects
Published in Journal of Intelligent Manufacturing, 2015
Research Overview
This research addresses complex scheduling optimization problems in manufacturing environments where learning effects occur at multiple levels - both job-specific and machine-specific learning phenomena are considered simultaneously.
Key Contributions
Problem Formulation:
- Develops mathematical models that capture dual learning effects: workers improve performance through job repetition (job-dependent learning) and machine operators gain efficiency through equipment familiarity (machine-dependent learning)
- Incorporates realistic manufacturing constraints where learning rates vary across different job types and machine configurations
Solution Methodology:
- Implements multiple heuristic algorithms to solve the computationally complex scheduling problems
- Compares performance of different algorithmic approaches including genetic algorithms, simulated annealing, and custom constructive heuristics
- Provides computational complexity analysis and performance benchmarking
Industrial Applications:
- Demonstrates practical applicability in manufacturing systems where both human and machine learning contribute to productivity improvements
- Shows significant improvements in makespan reduction and resource utilization when learning effects are properly modeled and optimized
Impact and Significance
Published in the Journal of Intelligent Manufacturing, this work bridges the gap between theoretical scheduling research and practical manufacturing optimization, providing tools that can be directly implemented in smart manufacturing systems where continuous improvement through learning is a key competitive advantage.
Recommended citation: Peng-Jen Lai and Hsien-Chung Wu. (2015). "Using heuristic algorithms to solve the scheduling problems with job-dependent and machine-dependent learning effects." Journal of Intelligent Manufacturing. Volume 26, Issue 4, pp 691-701.
Download Paper