Could someone juice up an Aibo so it could run faster?

If I had the time, energy and money to spare, and I was an electrical/mechanical/software engineering student, I would grab an Aibo, replace the motors, jack up the power, and see if I could make a really fun little robot.

Is this realistic? What is holding Sony back from making Aibo far more dynamic?

Another suggestion:

Sony Aibo Suped-Up Soccer Team Robocup 2003

Aibo is very cool and that it competes in Robocup in its very own division is very cool. Nonetheless, they should create a modified Aibo league to encourage hackers to juice up Aibo’s so they can run and thus stimulate a far more interesting game.

Let’s unleash some aggressiveness in the world of Robotic Dogs. Let those MIT kids and other engineering wiz’s see if they can juice up an Aibo so it can run down a field. I know it will probably crash and fall often…but then Aibo can just stand up.

Let hackers replace servo’s, moters, power packs, whatever, so we get far more interesting and entertaining robots. Sony can then steal any really good ideas and integrate them into a new Aibo.

Let’s see…

First, you’d want to increase the power output from the motors. The Aibo uses something like a dozen motors to walk, and they’re tiny little expensive precision-made things. Replacing all 12 motors with higher-powered models would require a lot of re-engineering, since they new motors probably wouldn’t fit where the old ones were. You might be able to get away with running the original motors at higher voltage for more power, at the risk of burning them out.

Secondly, you’ll have to replace the solid-state motor driver electronics with higher current and/or voltage parts. This will get expensive.

Thirdly, you’ll probably have to seriously modify the software that controls the walking motion, since ithe limbs will all be moving faster than before.

Fourthly, all these extra power demands are going to drain the interal batteries a lot faster, although for competition needs that’s probably not an issue.

None of this is impossible, but it won’t be easy. Sounds like an interesting project, if time and money consuming. It might almost be easier to scratch-build your own design.