Municipal solid waste collection: An effective data structure for solving the sectorization problem with local search methods

This paper deals with the crucial sectorization problems regarding household waste collection. Our purpose is to construct a fixed number of sectors which should be balanced with respect to daily total time for collection tasks. First, we propose a model and point out that our taking into account the number of connected components of each sector results in a complex large-scale combinatorial problem. Second, we present a new effective data structure, called the sectorization matrix, which allows sectors and economic updates to be suitably represented. Finally, considering two approximate versions of the original model, we advance that the initial numerical experiments indicate that basic local search methods using moves relying on this data structure can improve a feasible solution initially given by the planner, within reasonable time requirements.