Amber Krishan, Justin Blinder, and I have begun prototyping solar robots in phys comp. The current goal to is to create a kinetic object, powered using solar energy. The only problem is, solar cells are not the most efficient, reliable, or consistent power source, so it is quite a challenge. Walking/rolling robots have been done sooo many times in so many ways, that we decided a flying robot was the way to go.

"The Emu"
As I mentioned, solar energy is not the easiest to harness, and the key to ‘amplifying’ the power source is using large capacitors and a 1381S74U voltage detector (as well as several PNP and NPN transistors). The schematic for this 1381-based solar engine came from solarbotics. These components (particularly the caps) add a lot of weight to the structure. This translates to failure. We kept the structure as light as possible, using hollow plastics, as little perfboard as possible (in fact the first prototype used a “dead bug” configuration, soldering the legs of each component directly to each other), and thin flexible solar panels.

In the end, we got all four blades to spin (these are hacked vibration motors), but failed to take off due to excessive weight, and let’s be real: not enough power. It seems to me that building really small bots using a photo-popper circuit will be the only way for any real success with solar energy.
