Methods to solve the scheduling and routing problem to clean debris in post-disaster relief

One of the main challenges in post-disaster logistics is the removal of debris from impacted areas to initiate the recovery phase. We consider the problem of debris removal, focusing on the planning and coordination of a set of work-troops and a fleet of vehicles to remove debris and transport it to landfills. This problem, referred to as the Scheduling Routing Problem to Clean Debris (SRP-CD), is especially complex due to the massive volume of debris and the extended time horizon required to perform the cleaning in many real-world disasters. This raises significant challenges as it is a large scale optimisation problem. Existing methods, including metaheuristics, struggle to provide good quality solutions. We propose a mathematical model that emphasises the scheduling aspect, while implicitly managing routes by leveraging the repetition and the volume of debris, to perform full truckload transportation accordingly. In addition, we introduce a matheuristic in which vehicles are dynamically reassigned to further improve resolution. The methods presented in this study are designed to solve realistic large scale scenarios, like the Port-au-Prince Haïti 2010 earthquake case study. Proposed methods improve the best known results, reducing cleanup time for all scenarios while requiring a much shorter computational time.