This week, we focused on integration of each of our separate components together. The majority of the work we’re doing at this point is integration through pair programming. The part that I programmed this week was the control flow of each of the three threads (obstacle detection, SLAM/path planning, path execution) happening at the same time. For the end condition, we decided that the program would be complete when it is enclosed by walls on all sides and has mapped out the entire interior.

I also wrote some optimizations for speed and picture resolution. This included optimizing our search algorithm to not plan repeat destinations and learn from where obstacles are placed so it doesn’t run into them again. I also wrote a post-processing compression function that generates the final user map display.

Now, I am confident that we will have a real time project, but am slightly worried that the optimizations we find will not be enough to produce a useful and accurate result. This next and final week will be focused on making as many optimizations for the user as we can, and working on the final presentation which I will be presenting.