Mixed-Integer Linear Program for the Operational Harvest Scheduling Problem: An Application to the Thai Sugar Industry

Thai sugar industry is relatively unique due to the existence of local restrictions and practices, such as grower equity and sugarcane farm burning. All of which are interrelated and stipulated by different conflicting objectives of different key supply chain actors. In order to address these issues, a mixed-integer linear programming model for this so-called Operational Harvest Scheduling Problem (OHSP) is formulated and solved under four different aspects: (i) maximizing sugarcane input, (ii) maximizing grower profits, (iii) minimizing industrial yield loss, and (iv) minimizing environmental impact. Our preliminary computational results on 30 fictitious instances indicated that, regardless of the chosen objectives, the resulting sugarcane input did not differ much, with the maximal gap around 10% across four settings. Nonetheless, the other three objective values differed greatly and depended on the harvesting resource selection. More specifically, green sugarcane harvesting was the best in terms of both carbon emission and grower interests, while burnt sugarcane harvesting was the best in terms of extractable sugar amount. We also found that the number of harvesting resources largely affected the quality of the OHSP solutions.