If we're trying to introduce robotics, getting the robot adapted to the environment as it stands now is only one approach. Getting the environment to be suitable to the robots is at least as important. I wonder what other infrastructure we'll have to provide to robots... it seems like a lot people are working more universal solutions to charging up robots at they go along.
I keep wanting to take a small inflatable blimp and have it fly around. Infrastructure would then be a small landing pad with an electromagnet to hold the buoyant blimp up so the fans don't have to continue to run while charging. For charging i've been thinking of using the same tech as wireless charging pads for cell phones. This means I'd need somewhere the blimp could find as home to land on. As far as helium replacement I haven't figured out a sane solution yet. Combined with this rfid improvement then I'd possibly be able to finally have a computer that i can ask "Where the hell are my keys" when I'm running late for work in the morning.
You might be able to work off of a nexus 5 and the official charging pad as a starting point. They've got magnets (though permanent, not electromagnets) to hold the phone in place when charging.
Cool! I'm glad to see RFID and localization being merged together like this. I think this really shows that RFID is a much more suitable technology for this type of application than a computation-heavy CV algorithm.