Most of the robots I build are for teaching students about robotics, so they tend to be at the basic end of the spectrum. However, my project students build proof-of-concept robots to do all kinds of things – transport packages, carry your clubs around the golf course, simulate human breathing to calibrate a radiotherapy machine, etc.
The AI robots I’ve worked on are just simple robots – they’re really designed to learn about how AI can be used to control robot movement. I’ve supervised a couple of students using artificial neural networks (one type of AI) to make a two wheeled robot balance itself. That kind of project is perfect because it’s simple enough to get working, but really lets a student get a better understanding of AI in real engineering (rather than science fiction). AI is a really exciting are and I have no doubt at all that it will be really important in the future of robotics, especially as robots become more complex and as we ask them to do more complicated tasks.
Unfortunately, I don’t get to build robots as part of my day to day work.
It seems like a really interesting area to work in, and something I’d love to do at some point in my career.
AI looks really interesting, but I don’t think it’s quite at the level that’s portrayed to the general public just yet.
Comments