Every year, humanitarian organizations assign a sizable portion of their limited financial resources to procure, operate and maintain operating assets, without which service delivery would be nearly impossible. In this paper, using vehicles to represent operating assets, we identify policies for sizing and allocating operational capacity to minimize the expected deprivation costs in a humanitarian development context. First, we develop a stochastic dynamic programming model, and then an efficient heuristic policy that considers the interaction of asset purchasing and operating decisions when the budget is uncertain. Based on a data set provided by a large international organization, we estimate the parameters of our model to run numerical experiments. Results demonstrate the following: (i) Although budget uncertainty increases the expected deprivation costs and decreases capacity utilization, the negative impact of budget uncertainty is mitigated if budget savings between periods is allowed; (ii) a policy for minimizing the expected deprivation costs over time may avoid using all available assets in all periods; (iii) in situations in which the variations in the criticality of missions are large, both the expected deprivation costs and fleet utilization decrease; and (iv) in most situations, a centralized asset procurement model outperforms a decentralized model, not only in terms of logistic costs but also in minimizing the expected deprivation costs. Link to the article
KESHVARI, F., EFTEKHAR, M. and PAPIER, F. (2019). An Approach for Managing Operating Assets for Humanitarian Development Programs. Production and Operations Management, 28(8), pp. 2132-2151.