Optimal task sequencing and aborting in multi-attempt multi-task missions with a limited number of attempts

Mission abort policies have been investigated for both single-attempt and multi-attempt missions in the last decade. The existing models typically assumed a single task performed during the mission. However, a mission in practice may consist of multiple tasks (e.g., a surveillance mission consisting of multiple tasks with different routes). This paper advances the state of the art on aborting policies (AP) by modeling systems performing a mission with multiple tasks. Each task may be executed under a different environment and have a distinct AP based on the number of shocks experienced and on an operation time threshold. Each task may be attempted multiple times and the total number of attempts is limited by the available system resource. The operating en-vironments during the operation phase and the rescue phase of each task may also differ. The task-dependent AP and the execution sequence of multiple tasks are jointly modeled and optimized to minimize the expected mission losses (EML). The solution methodology encompasses a new recursive EML evaluation algorithm and the genetic algorithm-based optimization method. The proposed AP model and solution method are demonstrated using a case study of an unmanned aerial vehicle performing a five-task surveillance mission.

Automated summary

Mission abort policies for both single-attempt and multi-attempt missions have been investigated. This paper advances the state of the art by modeling systems performing a mission with multiple tasks. The proposed solution method is demonstrated using a case study of an unmanned aerial vehicle performing a five-task surveillance mission.