Research Project
Multi-Agent Multi-Task Planner
Description
The current approach in the IT method finds the best plan through the combined roadmap while assuming that a robot of the underlying robot-type roadmap is always available. When planning for a sequence of tasks this is not always the case. There are scenarios in which the optimal plan requires waiting for robots to be able to complete certain subtasks along the way, but a different suboptimal plan does not require waiting and thus becomes the more efficient plan give the constraint of robot availability.
Objective
The goal for this summer is to find an optimal plan for a sequence of tasks using a multi-agent/robot team that accounts for shared and limited resources by dynamically updating the calculated path as tasks and resources are modified.