Publisher's Synopsis
Incorporating human preferences and unique domain knowledge into scheduling processes enhances both efficiency and employee well-being. However, careful planning to ensure effective and employee-friendly use of human resources and their expertise is both essential and challenging. By aligning work schedules with individual preferences, employers can boost job satisfaction, reduce staff turnover, and enhance job attractiveness, which is critical amid today's skilled labor shortages. Additionally, accommodating preferences improves productivity, reduces workplace stress, and promotes better mental and physical health. Beyond preferences, human expertise is a valuable yet often underutilized resource in scheduling decisions. In complex domains, humans bring context-specific knowledge and years of experience that cannot be easily quantified or encoded in algorithms. By combining this tacit, domain-specific understanding with data-driven optimization and machine learning techniques, organizations can achieve more effective scheduling outcomes while making optimal use of limited human resources. This thesis presents innovative methods to harmonize human-centered and data-driven approaches, and demonstrates their effectiveness for railway crew scheduling and surgery scheduling. The resulting methods not only improve operational performance but also ensure more employee-friendly workforce management practices, with relevant implications for academia and industry.
