A Novel Approach for Production Planning and Productivity Improvements in Dynamic Production Systems using DES and Heuristic Methods

Publication Type:
Thesis
Issue Date:
2024
Full metadata record
The contemporary manufacturing landscape, characterized by global competition and the advent of Industry 4.0, necessitates a paradigm shift in production systems. Small and medium enterprises (SMEs) face the daunting task of surviving in this highly competitive environment marked by short product lifecycles. Even larger enterprises grapple with the challenges posed by increased connectivity and complexity resulting from Industry 4.0 tools. This complexity is further magnified in diverse production settings, such as the High-Mix Low-Volume (HMLV) and Low-Mix High-Volume (LMHV) environments, each presenting unique scheduling and variability challenges. This thesis addresses the pressing need for an advanced production planning approach capable of accommodating uncertainty and capturing the intricacies of different industrial settings. It identifies a knowledge gap in existing research, highlighting a lack of integrated approaches that combine heuristic optimization algorithms and discrete event simulation (DES) comprehensively. While heuristic algorithms often overlook real-world stochastic and dynamic elements, existing stochastic methods, such as stochastic programming and fuzzy programming, exhibit limitations in accuracy and manageability. The research focuses on developing a holistic framework that seamlessly integrates heuristic optimization algorithms and DES, providing decision-makers with a more realistic and comprehensive perspective on the challenges associated with production systems. The proposed approach aims to address multiple production challenges in stochastic and dynamic systems, offering adaptability across diverse industries and operations, whether manual or automated. Ultimately, the goal is to enhance overall productivity and streamline the planning processes in the ever-evolving landscape of modern manufacturing. This thesis was conducted based on three studies that capture different industrial settings: 1. Woolshed industry as an example of a heavily manual process, 2. Additive manufacturing as an example of (HMLV), 3. Assembly line balancing as an example of (LMHV). Research findings show the effectiveness of the proposed approach in diverse production settings. In the woolshed industry, challenges related to facility layout and resource planning are successfully addressed. In additive manufacturing, the integration of discrete event simulation and genetic algorithms reduces total production time, with notable advantages as the factory scales up. The study on assembly line balancing demonstrates a 15% improvement in throughput and resource utilization compared to heuristic methods in isolation in the low mix high volume industry.
Please use this identifier to cite or link to this item: