Dynamic Volunteer Staffing in Multicrop Gleaning Operations

• 成果类型: Article
• 署名作者: Ata, Baris; Lee, Deishin; Sonmez, Erkut
• 署名单位: University of Chicago; Western University (University of Western Ontario); University of Nebraska System; University of Nebraska Lincoln
• 刊物名称: OPERATIONS RESEARCH
• ISSN/ISSBN: 0030-364X
• DOI: 10.1287/opre.2018.1792
• 发表日期: 2019
• 页码: 295-314
• 关键词: heavy traffic analysis; Asymptotic Optimality; multiclass queue; random yield; myopic heuristics; parallel servers; channel subject; variable-yield; or/ms research; power-control

摘要

Gleaning programs organize volunteer gleaners to harvest a variety of leftover crops that are donated by farmers for the purpose of feeding food-insecure individuals. Thus, the gleaning process simultaneously reduces food waste and food insecurity. However, the operationalization of this process is challenging because gleaning relies on two uncertain sources of input: the food and labor supplies. The purpose of this paper is to help gleaning organizations increase the (value-weighted) volume of fresh food gleaned by better managing the uncertainties in the gleaning operation. We develop a model to capture the uncertainties in food and labor supplies and seek a dynamic volunteer-staffing policy that maximizes the payoff associated with the amount of food gleaned. The exact analysis of the staffing problem seems intractable. Therefore, we resort to an approximation in the heavy traffic regime. In that regime, we characterize the system dynamics of the multicrop gleaning operation and derive the optimal staffing policy in closed form. The optimal policy is a nested threshold policy that specifies the staffing level for each class of donation (i.e., a donation of a particular crop type and donation size). The policy depends on the number of available gleaners and the backlog of gleaning donations. A numerical study using data calibrated from a gleaning organization in the Boston area shows that the dynamic staffing policy we propose can recover approximately 10% of the volume lost when the gleaning organization uses a static policy. To achieve this improvement, no capital or major process changes would be required-only some small changes to the staffing level requests.