Bossa Nova is a company committed to bringing you a home robot in the very near future. Its latest model was announced at Robotbusiness. There's an SDK and you can start thinking up uses for it right now.
It is generally believed that the domestic robot market is the next big thing - when exactly 'next' is, is more of a problem. Bossa Nova has been trying to break into the big time robot market by first making robotic toys. This isn't a silly idea because the robotic toy market, being part of the general toy market, is a thriving marketplace.
Bossa Nova is now aiming to move from toys to bring you the equivalent of the Jetson's Rosie the robot, but its latest design, mObi isn't quite there yet, but it is an interesting design.
As soon as you look at it, it is clear that this is a robot that balances on a ball to get about. This sort of robot locomotion, ballbot, was invented at Carnegie Mellon's Robotics Institute. The ballbot is capable of moving in any direction and of course turning on the spot. The only problem is that it is going to fall over as soon as the power is off. The solution to this problem is to deploy three-legged stand. It is claimed that the ballbot allows mObi to move around a "human" environment better than a wheeled robot. Of course this means that mObi has the "Dalek" problem, i.e. no ability to cope with steps or stairs.
The hardware in mObi is interesting . An Intel processor runs either Windows or ROS (Robot Operating System) Its main sensor is a PrimeSense 3D camera, i.e. a Kinect-like device. Apart from this it isn't clear what other facilities will be available.
So what can you use a robot like this for?
Without any manipulators, its functions are limited to telepresence or surveillance duties. Bossa Nova intends mObi to be used for research in human-robot interaction studies but this is probably aiming a bit too low. There will soon be an SDK and more information. The plan is for mObi to be availalbe next year.
Results of the first ever Winograd Schema Challenge were unveiled recently at the International Joint Conference on Artificial Intelligence (IJCAI-2016). The low scores reflect how far there is [ ... ]