Is it far fetched to partner with TSLA for some basic AI IP. And apply it to teach the 360 how to map a home, one would think this application would be a breeze for AI. I feel all the heavily lifting is already done with the algorithm. Is this far fetched or the robots vision, cpu not capable enough? My 360 is being very inefficient, where it could do long straight sweeping runs down a hallway and back, for example it is  vacuuming the hallway horizontally in a square motion which causes it to make significantly more turns and bumps and reverses etc. 

thanks