We propose a binary linear programming model to solve a minimization problem of working time for a harvester. This machine is used by an agricultural cooperative which has to process a big set of smallholdings.
The complexity of the model is a consequence of the great number of involved variables and constraints. The variables use triple index, related with the number of smallholdings and the number of invoiced periods of time.
The main constraints appear because processing and displacement times are invoiced, the portions of every owner should be processed as a block, the time required for working every area is given and every owner has a proposal about the period of time in which he needs that the machine starts the work in his land.
We define two heuristic algorithms to approximate the solution and we implement them. One of the algorithms is based on local search methods and the other one on classical tabu search heuristics.
Finally, we obtain computational results by considering a set of data taken of a real case.