Formula Pi Winter 2017/18 - Update 3

So the last challenge round of the season didn’t quite go to plan - Team Limilo now sits back in 13th place, after yet another zero point haul from an obstacle avoidance based heat.

Simulation testing had gone pretty well, but this was a reminder that nothing can subtitute for running code in the real world. I’d taken a chance on reducing the threshold for triggering an overtake which didn’t pay off at all. The robot hit the first obstacle and took an age to finish it’s first lap. It’s opponent wasn’t uncatchable at that point, but on the second lap the robot hit yet another obstacle, this time triggering a complete reboot. Unfortunately this put it into ‘waiting for the start sequence mode’ and that was the end of the race as far as Team Limilo were concerned.

So, a couple of takeaways from that. The boot sequence should really check for the existence of race logs, indicating that the race had already started (therefore bypassing the start sequence holding pattern). Secondly, just tweaking parameters for the avoidance code isn’t enough - there needs to be a whole rethink.

One last race remains. The top 5 final is now beyond reach, but a top 10 place would be a decent return for a first season I think, so testing is geared towards getting round as fast as possible. There’s also the lap record to aim for - an impressive 13 and a bit seconds. Claiming that would be a real prize.

The last challenge heat is available here on YouTube.