Abstract
This paper presents a mixed integer linear programming (MILP) model to aid in planning a large, hierarchically structured workforce. The workforce (eg, all Army commissioned officers) is classified by occupational group and rank and modelled in yearly intervals. Personnel and their movements are modelled as stocks and flows, subject to constraints representing employment conditions and resource limitations. The task is to estimate personnel numbers and flows so as to minimize a composite cost function. The model’s detailed fidelity to actual conditions, and hence its value in practice, clearly exceed previous efforts in the workforce planning domain. To overcome computational challenges that arose when an MILP solver was applied directly, the authors developed an iterative solution approach, which has yielded an attractive combination of solution quality and computational performance.
Acknowledgements
David Oliver of the Australian Department of Defence, initiated and supported this research. The authors’ colleague, Andreas Ernst, provided valuable ideas and advice.