Modeling a fresh fruit harvesting and distribution problem in resilient agribusiness supply chains

The paper presents an integrated modeling approach to support decision–making in a three–echelon agribusiness supply chain to harvest, process and distribute fresh fruits to different markets under disruptive scenarios. The harvest problem includes a delineation of the field into management zones and a harvest scheduling based on optimal harvest time windows for each management zone depending on the quality of the fruit. The processing problem consists of extending the shelf–life of the fruit, a process widely used in the fruit export sector to delay the oxidation of fruits and vegetables. The distribution problem includes possible disruptions in ports and transportation, which is critically important for fresh products. The proposed methodology considers a mixed–integer two–stage stochastic optimization model with the goal of making a resilient planning framework for fresh fruit supply chains. Moreover, the risk–neutral model is extended by applying two quantitative risk management formulations: one based on the Conditional Value–at–Risk (CVaR) measure and another aimed at minimizing the absolute negative deviations with respect to the distribution average (MOTAD). The Value of the Stochastic Solution (VSS) demonstrates that incorporating uncertainty allows the model to find better solutions. Besides, integrating risk measures in the two–stage stochastic program forces it to adopt more conservative operational strategies to handle disruptive scenarios in the supply chain. The CVaR approach makes better operational decisions aimed at shipping fresh fruit with higher remaining shelf–life levels to key destinations in specific adverse scenarios, whereas the MOTAD method prioritizes achieving overall stability in meeting the required shelf–life across all scenarios.