A state-of-the-art survey on multi-scenario scheduling

In classical deterministic scheduling models, it is common to assume that each of the parameters has only a single predefined value. However, in many applications this is not the case, and there is a certain level of uncertainty in the values of the parameters. The multi-scenario approach is one of the most common approaches to model uncertainty. In this approach, each scenario defines a different possible realization of the parameters and the scheduler has to decide upon the exact schedule prior to knowing the exact scenario realization. Different scenarios may arise for a large number of reasons, such as: (i) the assignment of different workers to the processing operations, where the identity of the worker is beyond the scheduler's control; (ii) unpredictable changes in weather conditions; or (iii) a possible strike at the port, which may affect job release dates. The purpose of this paper is to offer a unified framework for scheduling in a multi-scenario environment by presenting an up-to-date survey of results in this field. We also included some new results, presented here for the first time, and suggest many challenges for future research.