Graphical scheduling heuristics for complex task environments

This paper presents a heuristic approach to scheduling the execution of interdependent tasks in a complex environment. In this context, tasks require the assignment of a set of assets within a specified opportunity window. The tasks are geographically distributed and require the movement of assets prior to processing. Each asset has multiple resources and is generally combined to meet the requirements of a given task. In this paper, a graphical task model is introduced that facilitates hierarchical, sequential, and parallel task relationships. The schedule is determined by sequentially assigning assets to tasks using a greedy heuristic. Each assignment corresponds to a subproblem, where a set of assets is selected to process a single task. The performance of this heuristic is improved via a rollout algorithm.