There has been a lot of discussion recently by Intel's CTO (Justin Rattner) about some really compelling future technologies: wireless power and programmable matter (made of catoms). Of course, the programmable matter (catoms) he is discussing are basically robots operating as a swarm. Wouldn't it be neat to see the swarms actually powered wirelessly? While Intel has thus far worked on the two technologies disjointly, work presented by myself at ICRA 2008 is addressing the intersection -- wirelessly powering a swarm of robots (publication here).
Wireless power transfer is nothing new; it has been discussed since Tesla's patent in 1900 entitled "Apparatus for Transmission of Electrical Energy" (USPTO #649,621). However, as the technology matures, it will be interesting to see what myriad applications arise.
For example, Intel's system is capable of transmitting up to 60 Watts (the demo of lighting an incandescent bulb) around 75% efficiency. They hope to one day use the system to remotely charge laptops.
Meanwhile for robotics, wireless power has the capability to transform research and applications involving swarms of small (mini/micro) robots. When dealing with robot swarms numbering in the hundreds or even just a dozen, tethering is impractical and changing batteries is cumbersome. By way of an example, consider the battery-powered robot swarm by Caprari from EPFL in Switzerland, where they have ~100 robots operating simultaneously.
Hooking up 100 robots for battery charging does not sound like such a fun prospect -- more importantly, it is a significant research impediment. Instead, it is possible to use inductively-coupled wireless power, in a manner similar to Low-Frequency (125kHz) RFID, to power the swarm or just to perform simultaneous, contact-less battery recharging. This is the goal of the work presented by Deyle (myself) and Reynolds in Surface based wireless power transmission and bidirectional communication for autonomous robot swarms. What's great about this work is that the fundamental design is straight-forward, and can be prototyped using discrete components and basic microcontrollers, making the technology available to researchers and hobbyists alike.
The surface is outfitted with a transmit coil. This coil operates in resonance to increase the circulating current and thus increase the magnetic flux -- providing increased power density on the surface.
The magnetic flux is coupled into a receive coil on the bottom of the robot to provide power.
Now all of the robots can either operate simultaneously on the surface, or they can sit still and recharge on-board energy storage (such as a large capacitor or battery)! Here is what it looks like in operation (along with a wirelessly-powered LED).
For more extensive details, I recommend checking out the publication, here.