They are the one company that makes me believe that robotics may
someday actually be used outside of a controled enviroment. When I
first saw Big Dog survive that kick, I know my mouth hung open for a
while. I anxiously await their next generation of robots.
I've been expecting that for almost twenty years. Raibert did some
great work in the 1980s at the MIT Leg Lab, but then he left MIT to
form a startup company, and it was many years before Boston Dynamics
started doing interesting dynamic robots again.
I did some work in that area back in the early 1990s, but we didn't
have enough compute power back then to get anywhere near real time.
Now we do. I have a patent in this area. (#5,644,204)
How it survives a kick is well understood. The robot has a full
6DOF inertial system (which isn't that expensive any more), and computes
the zero moment point (look it up). Then it tries to get at least three
legs placed in a triangle that covers the zero moment point. If it can
do that, it won't fall.
Don't forget RHex: http://bostondynamics.com/content/sec.php?section=RHex
One cool thing about Boston Dynamics is their close connection to the
academic world. With strong academic roots (being an offshoot of the
MIT leg lab), Boston Dynamics still remains very closely connected to
acadamia. If you search Google Scholar - you'll find many papers that
have surfaced from within the company. Also, both RiSE and LittleDog
are being worked on in conjunction with a number of other
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.