The straight line speed test was causing some concern during testing using the distance sensors. So with less than a week to go we decided to rip up all the code and switch to line following using the camera.
In our first year competing at Pi Wars we did fairly well at the line following course but not the second year. We were confident though that we knew what was causing the issues and had ideas on how to make it better. A long week of rewriting the code and testing every night followed. We must have rewritten the whole module at least 4 times that week, mostly during my lunch hour at work and then we would spend all evening testing it. At 10pm the day before Pi Wars we were finally happy with what we had done for the challenge and at least expected it to be able to get down the course if not very quickly.
On the day things started well as it trundled off down the course until it got to a part of the line that was very dirty (it had a boot print on it!!) and had a gap in it. Well, the bot did not like this and decided it was going to reverse back towards the start of the course.
Things just went downhill from there really and that was as good as it got. This was easily our worst event even though on the face of it the challenge is the most straight forward of the autonomous challenges. Better luck next time I guess.