It’s a tough question – we are working hard on figuring this out. Conventionally what is done is that we use normal PCs as the robot’s brains. We then connect these computers to sensors and then write algorithms that can do some of the tasks we want them to do. Recently it’s become really popular to use machine learning – these are basically algorithms that the more you run them, the better they perform.
Personally I don’t think PCs are the way to go. I think we need a new type of computer – one who’s architecture closer resembles an animals brain. It’s a very interesting problem…
Comments