Scheduling deliveries under uncertainty

Quite often transportation companies face two types of jobs, ones which they can plan themselves and ones which have to be done on call. In this paper we study the scheduling of these jobs, while we assume that job durations are known beforehand as well as windows in which the jobs need to be done. We develop several heuristics to solve the problem at hand. The most successful are based on defining an appropriate buffer. The methods are assessed in extensive experiments on two aspects, viz. efficiency, in the sense that they carry out many jobs and certainty, in the sense that they provide information beforehand about which jobs they will execute.