For those of you who aren’t Robot Geeks and missed DARPA’s Virtual Robotics Challenge mid this year, don’t worry! Only one day ago the DARPA Robotics Challenge Live Trials have just completed
(Spoiler Alert – I’ll reveal the winner later on :)
In case you’ve never heard of the DARPA Robotic Challenge and are wondering it’s significance, they had another one called the Grand Challenge for Autonomous Vehicles. It started in 2004 and no vehicle could complete it! They then scheduled another for 2005 which Stanford won with a tricked-out VW Touareg R5 called Stanley. This wasn’t the end though…
We then moved on to the Urban Challenge, which was won by Boss, a Chevy Tahoe with Stanford getting second place with Junior. Why is that significant? Because Sebastian Thrun (in the photo above) is the Director of the Stanford AI laboratory and also a Google Engineer. Yes, the Google Car is based on Junior, a modified Volkwagen Passat Wagon which came second in the Urban Challenge. Now back to the present and the Robotics Challenge…
You can find out more at the actual DARPA Robotics Challenge site or from the DARPA YouTube Channel which has a lot of content! If you want a bit more background, first watch Robotic Expectations & The Challenge History. The teams are all over the place, so here’s a slightly more structured curation of teams that had their own unique robots:
|Yes, that’s a female robot!||Uses flexible “muscles”|
|A mostly 3-D Printed robot||Which is based on a chimp|
|Essentially built from one component!||A clever multi-limbed robot|
|A lightweight robot||One of the few Japanese designed robots|
|An Open Architecture Robot!!!|
In the end though, the top 8 (who will receive continued funding) were:
- Schaft (27)
- IHMC Robotics (20)
- Tartan Rescue (18)
- MIT (16)
- Robosimian (14)
- TRACLabs (11)
- WRECS (11)
- TROOPER (9)
I’ve bolded and underlined the ones based on Atlas (which are 1/2). Why is this significant? Because Google just bought Boston Dynamics, who manufacture Atlas!!! What’s more, Google also owns Schaft who won! Monopoly anyone?…
As we saw from the Vehicles, it doesn’t matter whether a Google team actually wins. One thing you can guarantee – Google will be making robots in a few years which will be commercially available by the end of this decade. Given Google’s “record” – i.e. blatantly violating peoples privacy and who knows what else, I’m not sure how good this really is…
Yes, I know they’ve feigned disgust at various NSA revelations, but remember that their chairman Eric Schmidt, once said “If you have something that you don’t want anyone to know, maybe you shouldn’t be doing it in the first place”
I’m still waiting for him to mail me his social security number and bank account details, because hey, he’s obviously an upstanding guy who has nothing to fear… ;-) How would you feel about having a Google Spy (oops Robot) in your home?
One question remains though – where are Honda in all this? Ironically, the Japanese did win this challenge, but using technology which is pre ASIMO. I think HONDA are quite rightly keeping to themselves as what do they really need to prove? If you look at any recent ASIMO information you can see that it’s way beyond where all the DARPA people are, which is not to say they won’t catch up, but who knows what’s in HONDA’s labs?… The last information we have is from 2011 which is pretty awesome!
Also, they don’t really need to participate in the “DARPA Challenge” as they’re already working on a Disaster Response Robot Based on ASIMO and are using ASIMO to Act as an Autonomous Explaining Robot – beat that DARPA door opening droids! ;-)
My hopes and predictions (since we’re getting near New Year) for the future on Robots?
Google will obviously have a number of models with military and civilian applications which will become a major profit centre for them (thanks military industrial complex :) – not a surprise considering all the “hard AI” work they’re doing with Kurzweil – stay tuned next decade…
Honda will enter with ASIMO and buddies around the same time. They’ll probably be more expensive but will be more consumer friendly and secure (i.e. they (hopefully) won’t be spying on you).
Open Source will have something, and Hubo definitely looks like a good start – I like the concept that I could own a robot that I can trust and improve :-)