Stanford Robotics Club demoes onstage at TC Sessions: Robotics+AI.
TechCrunch is a leading technology media property, dedicated to obsessively profiling startups, reviewing new Internet products, and breaking tech news.
Subscribe to TechCrunch today:
[Music] How you doing great I thought that we Could just start if you could you just Introduce your little buddy in your hand Yeah this is the newest robot that we’ve Been working on It’s called Stanford pupper it’s a Low-cost open-source quadruped it that We hope lowers the barrier to entry to Robotics I can’t help but notice that You put googly eyes on the front of it Oh yeah is that to humanize it or does That do those have a like a function I’m very functional and making it more Cute making it more cute so that nobody Treats it badly exactly okay can we show Them absolutely what you did here it’s a Cute it’s a cute guy can you guys see so Stanford pupper is twelve degrees of Freedom so that means they can go Forward backwards side to side and Turning in place and can do kind of cute Sitting and Orient’s thing like that we Call that the rock the hip all right oh Yeah this one is the real rocking the Hip show them the sneak walk Oh sneaky Mode all right it’s like you never even Knew it was coming right And how about a jump Nathan can we do One of those oh yeah Oh I mean round of applause right very Good can you tell people like I don’t Immediately see in a commercial Application although I’m sure the people
In this room are like I need to go Acquire Nathan and sell this dog robot But what is the why are you guys doing This in the first place this is the Academic venture yes why we’re a student Robotics team at Stanford So we’re originally not associated with Any lab and we were just inspired by Quadrupeds and other legged robots that We’re seeing out there like Boston Dynamics I’m from the Cod lab from University of Pennsylvania but we we Really just want to play and I have one In our own hands and we didn’t have a Way to do that they were very expensive So we started out just like how can we Make one on our own budget and so that Led to using inexpensive parts while Finding the most powerful motors Powerful actuators and doing the Software itself so for this robot it’s Continuation of our original robot Stanford taco and we’re just trying to Make it as accessible as possible to People who are interested in robotics a Lot of people are interested in Stanford Bago but you know it’s kind of difficult To build it cost $3,000 so this one cost $600 and it’s quite easy to build you Can build in about eight to ten hours And so it’s our most inexpensive most Accessible robot yet and is there a Place for people in this room maybe Could go and find out how to build their
Own dog you I was a last night getting The website ready it’s Stanford seeing Robotics org slash robots and if you Want to build any of our robots you can Check out that website it’s very Generous of you Do you have more robots to show us you Want to bring some out or not have five More robots to show so Aaron over here Co-lead of the robotics group I will Show our original robot Stanford dog oh And yes this is our original robot dog Oh we presented to him last year at Accra in Montreal Doggo holds the record for highest Jumping quadruped ed at over a meter we Do that under laboratory settings but he Also holds the record of all robots for Highest specific vertical agility which Is sort of how high you can get how fast Yeah we do like a medium high jump I Know you won’t go a full meter but like Can we push the limits this is the Highest we’re hopping right now but I do Have another trick I can show you so That was a firm no then that that was a No but I got you we can do a back Handspring That’s cool I mean what was like Mobility these obviously you’re you’re Very experienced and seasoned robot Assists what was like the biggest Challenge right did you guys like run Into I know you were working really hard
To make it low-cost and affordable I’m Sure that gave you some big decisions in Terms of do we want X or Y like what was The big challenge or the big I think in All of robotics But what really like stood out to me While doing this is just the sheer Number of iterations it takes to get Anything to work the way you want it to I mean it took us nearly a year and a Half to get our first robot doc oh Really working to a point where we could Show it off and then even then we Weren’t content with where it was at and Wanted to keep working to improve things And you know slowly did that through a Bunch of iterations I also want to ask Like you guys are clearly the future You know when it when you’re talking About I made a bit little joke about the Googly eyes but like in reality when You’re kind of balancing do you give it Human or at least familiar organic Features like aesthetically and do you Have to decide between the two do you Guys have a stance really on whether Robots should look and feel like Something that’s familiar or should they Be more like of the Roomba persuasion Where it’s just a robot it doesn’t need To look like anything but a robot right Well we specifically avoided trying to Make them look just like dogs I think When you’re on the uncanny valley that
Would be a little bit weird but other Than that we’ve been focusing on the Fuel like you were saying does it feel Like something that could be alive and From there that kind of builds a sense Of attachment as I think just walking Smoothly being able to like look around At you googly tiser to something goo as They’re just something we added last Minute but the feel is what we’ve been Focusing on and I think from there the Anthropomorphic aesthetic comes out of It can I Drive the Doge absolutely okay I’m gonna hand this to you they’re top Secret yeah well I think now could be a Good time to bring out our four other Robots How’d about that yeah absolutely so we Have three more peppers coming out like I said they’re low-cost and so you’re Like why not we can make five in total This larger robot here is for research It has torque controllable actuators a Very agile robot we think it can jump About two meters high but we haven’t Tested that one out yet and this one is For research especially in our lab where We’re working on different types of Trajectory optimization to really use New powerful embedded computers to get The most out of robots and on the on Your guys’s right we have what we call Smart pepper this hood includes a Nvidia Jetson and then some cam depth cameras
And so for instance right now we have it Running an obstacle avoidance where if It gets too close to any one object you Can kind of like feel it and it’ll stop Going towards that I have one final Question for both of you well actually Have a question for the audience who Would be interested in hiring these guys A few of you yeah for sure what do you Guys want to do actually you’re both so Young like what do you what do you want To do with your lives and what are you Looking to go accom So you’ve got some talent here like what What are you looking for I mean I I’m Enjoying the journey right now and uh I’m still excited to see where it’ll Take me I’m lucky enough to still have At least one more year of school and We’ve actually started working with one Of the professors on campus to really Bring cutting-edge controls to this Robot in particular and that’s got me Interested in research and so I may be Looking to get into that field and more Immediately a masters or PhD cool Nathan What’s next for you yes I’m a senior at Stanford by sticking around for a year Doing a master’s program but after that I’m not quite sure a graduate program Perhaps like Aaron but really I’m Interested in industry and perhaps Bridging robotics and education and Allowing kids k12 undergraduates to get
Their hands as soon as possible on Interesting robots round of applause for These guys right awesome job Thank you so much you guys can find a Way to get your robots off stage I won’t Step on any of them they’re the cutest Little things just like little bugs Running around