Distribution planning problem of a supply chain of perishable products under disruptions and demand stochasticity

Purpose - The emergence of risk in today's business environment is affecting every managerial decision, majorly due to globalization, disruptions, poor infrastructure, forecasting errors and different uncertainties. The impact of such disruptive events is significantly high for perishable items due to their susceptibility toward economic loss. This paper aims to design and address an operational planning problem of a perishable food supply chain (SC). Design/methodology/approach - The proposed model considers the simultaneous effect of disruption, random demand and deterioration of food items on business objectives under constrained conditions. The study describes this situation using a mixed-integer nonlinear program with a piecewise approximation algorithm. The proposed algorithm is easy to implement and competitive to handle stationary as well as nonstationary random variables in place of scenario techniques. The mathematical model includes a real-life case study from a kiwi fruit distribution industry. Findings - The study quantifies the performance of SC in terms of SC cost and fill rate. Additionally, it investigates the effects of disruption due to suppliers, transport losses, product perishability and demand stochasticity. The model incorporates an incentive-based strategy to provide cost-cutting in the existing business plan considering the effect of deterioration. The study performs sensitivity analysis to show various “what-if” situations and derives implications for managerial insights. Originality/value - The study contributes to the scant literature of quantitative modeling of food SC. The research work is original as it integrates a stochastic (uncertain) nature of SC simultaneously coupled with the effect of disruption, transport losses and product perishability. It incorporates proactive planning strategies to minimize the disruption impact and the concept of incremental quantity discounts on lot sizes at a destination node.