Tuesday, July 27, 2010

Robot Maze Solver

It's been a long time I didn't write blog already. Why? Simple, because I'm busy and not much stuff to write. New zealand is getting boring and boring.. Luckily I gonna graduate next year, or else, I will have to stay here suffering from : no-Kopitiam, no-Makan, no-shopping mall, no-cheap movie ticket, no-RedBox, no-cheap food and no-Entertainment..

Today is the D-day of my group project to go into action! Line Follower Robot or Robot Maze Solver to be more 'cool'..After 2 months of planning, sketching, modeling, simulating, testing and troubleshooting, finally our robot went to the race day. Looking back at the process of building the robot, the hardest part for me would be the time when we were rushing for a report of the robot design. Combining the entire C codes and circuit design, it came out to be 40+ pages. Shocked but it was really hard. There was so much stuff that needed to be explained and fine-tuned.

PCB Design



Circuit Design



But at last, it did complete the maze but with some glitches. Below is video showing the first round of the race where the robot didn't actually complete the maze, but gone haywire.
video


We were given one hour to do some troubleshooting. Problems we faced were actually not that serious. The power supply wire came off which was bad. and we deleted a poart of the code to eliminate some probability function we used which seemed not working in real situation. Video below shows the complete one. It was timed at 1.30 minute for the race which was not that good in terms of ranking because of the algorithm we used to save memory in the processor. But in overall, we were quite satisfied.. Yay!
video