The ultimate purpose of the system_controls subteam is to figure out where the robot should move and execute on that knowledge; the perception is handled by the onboard sensors.
In pursuit of this goal, a strategy is needed to coming up with a path planning algorithm that is both effective and robust while being understandable. Most commonly, algorithms view a costmap as a graph with nodes representing areas on the map, and the best way to do this can be found with something like Dijkstra's or the A* algorithm. These may be overkill, and harder to debug, for our applications; successively driving in a straight line and turning may be a viable alternative.
The ultimate purpose of the system_controls subteam is to figure out where the robot should move and execute on that knowledge; the perception is handled by the onboard sensors.
In pursuit of this goal, a strategy is needed to coming up with a path planning algorithm that is both effective and robust while being understandable. Most commonly, algorithms view a costmap as a graph with nodes representing areas on the map, and the best way to do this can be found with something like Dijkstra's or the A* algorithm. These may be overkill, and harder to debug, for our applications; successively driving in a straight line and turning may be a viable alternative.