Local mapping + planning investigation

[b1n-ch1kn]

Few thoughts on this

• Would potentially replace fusion node (https://github.com/QUT-Motorsport/QUTMS_Driverless/issues/167)
• SLAM tracks lidar cones, vision maps colours onto them in the same way
• Local map to track only things in cars FOV
• As car passes pairs of cones, they could be popped off in order - then we have an order for track boundaries without computation
• Could also get a local midline to follow on first lap - which gets sent to the controller for a local look ahead waypoint

[b1n-ch1kn]

Delaunay triangulation. https://blogs.mathworks.com/student-lounge/2022/10/03/path-planning-for-formula-student-driverless-cars-using-delaunay-triangulation/?source=15571&s_eid=psm_15571

KA RaceIng path planning (3:38) https://www.youtube.com/watch?v=-NvZdvD-3wM&list=PLfkYTifHTw-foWXHFhT95gNP3sOostBX8&index=7

Possible approach is to create an ordered list of cones (x,y,col) as the car passes them on the map. It would 'pop' off mapped cones if the position of cones with respect to car is <0 and add them to the ordered list. Delaunay triangulation and centreline calculation on cone ordered list.

Testing: Check latency of map update + evalute path planning speed - will we need to update path every map update?