Workforce planning is an important process that: (1) enables organizations to determine the most efficient workforce composition, and (2) provides a foundation to recruit and/or reorganize the workforce to achieve organizational goals. An effective workforce plan has the right number of workers with the right skills for the right tasks at the right time. Unfortunately, simultaneously determining the workforce allocation - the number of workers with each skill set available during the horizon - and the workforce utilization - the sequence of tasks scheduled during the horizon to meet customer demand - is not a trivial task. In workforce planning problems one decision affects many others, resulting in many inter-connected, complicated constraints. Applying traditional techniques to solve workforce planning problems can lead to tremendous challenges. For example, the use of mathematical programming techniques can be hampered by non-linearities and weak linear programming relaxations.

This dissertation presents new mathematical models and solution techniques to accurately model and efficiently solve workforce planning problems. The work in this dissertation is motivated by existing problems in the service and manufacturing industries. Computational results based on data from a major automotive manufacturer demonstrate how the models and algorithms developed provide high-quality, realistic workforce plans in reasonable run times. This research could be extended to address other workforce planning problems in the manufacturing and service industries. In addition, the algorithms and models created can be applied to other resource allocation and utilization problems.