The video shows a multitude of iRobot products in action, including lots of PackBots with various payloads (including LiDAR), a PackBot deploying out of an airplane (awesome!), footage of the SeaGlider and Ranger submarine bots, and some kind of robotic boat. It’s a nice reminder that whatever we might think of the Roomba (which is at risk of being developmentally stagnant), iRobot’s still burning the midnight oil when it comes to military robots. And it makes sense, too, since (and I know I belabor this a lot), that’s where the money is.

Oh, and there’s also some kind of weird shooting thing at about 38 seconds in… I’m not sure what it is, some kind of weapon perhaps?

VIA [ RSN ]

After yesterday’s news that Pleo may get resurrected to some extent, it might be worth having a look back at just what happened to Ugobe. As shockingly entertaining as it might have been, Pleos weren’t systematically destroyed by Combots… Rather, it was a more banal combination of factors. Ugobe’s chief technology officer, John Sosoka, gave a lecture at Stanford a few weeks ago entitled “The Rise and Fall of a Companion Robot: Lessons Learned from Pleo.” You can watch it in its entirety here.

The talk is 40 minutes long followed by questions, but here’s the nutshell… Firstly, the economic collapse was obviously the main problem. Pleo is a luxury item, and an expensive one at that. But that’s the other part of it: nobody, Ugobe included, was really sure what they had with Pleo. It was kind of a toy, kind of a pet, and kind of a robot. Any way they tried to spin it, it was hard to market, especially for the amount that it cost.

Speaking of, Pleo is still available on Amazon, and the price has dropped to about $150. That’s only $10 more than Pleo actually costs to manufacture (the wholesale price was $195), according to Sosoka, so it’s a pretty good deal.

[ Stanford ] VIA [ GoRobotics ]

We’ve been waiting a year (a year!) for RoboGames, and this weekend, it’s all happening. Men and women, boys and girls from more than 20 countries will be competing in over 70 different events… There’s the combat (from micro-sized to 340 pound) of course, as well as the humanoid events, but I’m particularly looking forward to autonomous competitions and of course the big new event this year: Mech Warfare. We’ll be ringside, taking video. It’s gonna be AWESOME.

For up to the minute pictures and info, you can follow BotJunkie on Twitter. I’ll be Twittering from the event all weekend, and we’ll have oodles of pictures and video here on the blog by Monday. Unless I get killed by a rampaging combat robot. Then it might take longer.

The action starts this Friday and runs though Sunday at Fort Mason in San Francisco; buy tickets here… Or if you’re poor like me and you want to get in for free, volunteers are always needed. If you’re interested in that (it’s fun!), email me (evan@botjunkie.com) and we’ll work something out.

[ RoboGames ]

MIT’s Nexi, who we first met over a year ago, is now apparently qualified to captain US Navy warships, and not only that, but come to the rescue of crew members in peril. This video is meant to illustrate how different robots can work together to complete an objective with minimal human oversight, and even take a little bit of initiative on their own, like saving that extra human. Lucky us. The quadrotor pulled a pretty impressive trick, navigating in such a small area… Although I’m not sure something with four spinning blades would be the best thing to lead people down a low, narrow passage that’s also on fire.

[ Nexi ]
[ MIT LabCast ]

The new owners of Ugobe’s intellectual property, a company called Jetta, have pledged to resume production of the Pleo: “The company is firmly committed to re-launching Pleo and continuing the line including accessories such as the vital battery and charger components.” This is a pretty easy thing for Jetta to do, since they’re the ones who were manufacturing Pleo for Ugobe in the first place… They’ve got all the necessary infrastructure, and now that they own the IP, I guess they’ll just fire up the machines again and start cranking ‘em out.

Don’t get me wrong, this is certainly good news, I was just vaguely hoping that someone would buy the Ugobe IP and use it to make Pleo 2.0. Instead, Pleo belongs to a manufacturer, as opposed to a developer. We’ll just have to wait and see what comes of it, as there’s no information yet on a timeframe.

[ Jetta ] VIA [ RobotsRule ]

Rather than develop a robot that’s programmed to solve problems like a human, researchers at the AnimatLab in France are starting with a much simpler (and some might say more effective) problem solver: a rat. Psikharpax (I have no clue how to pronounce that, so don’t even ask) is a robot that not only looks like a rat, but is designed to operate using nearly all of the same sensory organs as a rat, including a set of sensitive whiskers. The software that controls the roborat is also designed to mimic real rats, relying on simple sensor inputs to dynamically build mental maps of complex environments.

So far, Psikharpax has eyes, ears, 32 whiskers per side, and (in a minor departure from its biological counterpart) battery powered wheels. A nose is in the works, and eventually Psikharpax will be able to sit up and grasp things with a set of front paws. At 50 centimeters long, Psikharpax will ultimately be fully autonomous, able to explore new environments and use learning behaviors to develop foraging techniques. And in case you were wondering (I know I was), “Psikharpax” was the King of the rats in the Batrachomyomachy, a parody of the Iliad written in Greek verses and (falsely) attributed to Homer. The name means “crumb robber”.

Crumbs, nothing. Robotic black plague, here we come.

[ AnimatLab ] VIA [ Physorg ]

This is a talk from Maker Faire by Rodney Brooks, of MIT and iRobot, on the “robotics revolution” that we keep hearing about. As you might expect from a guy who runs a few robotics companies, Brooks is extremely optimistic about the future of robots, comparing them to the evolution of the personal computer. The metaphor here is a very common one in the robot world: robots right now are like PCs were, 30 years ago. In other words, just wait a little bit, and robots will transform our world… Here, look at some mind-blowing graphs showing exponential progress!

Generally, I buy into this idea. I think robots have a promising (and exponential) future. But realistically, it seems as though we’ve got a little ways to go yet. There aren’t any robots out there that are like the PC was when it took off: something reasonably cheap, reasonably easy to use and understand, and capable of executing many different tasks (specialized maybe, but useful) in a reliable manner. Everybody is sort of still waiting for someone to invent that killer combination of software and hardware that will prove that robots are the future, and I’m not sure what it’s going to be. What my guess is, is that it’s going to be one of two things: a robot that does something very specific and important spectacularly well, or a robot that does many different things dependably decently. In the near term, it’s more likely to be the former.

From a different perspective, the robot/computer metaphor isn’t really appropriate. PCs specifically (and computers in general) found a niche because they kick us human’s asses at number crunching, hands down. We can’t do what computers do, so of course they’re going to be important to our development as a society. And robots are already successful and well established in the same way, at doing things that humans can’t or won’t do. Industrial work and bomb disposal are just a few examples. Problem is, humans can do a lot of things really, really well, and robots have to find some way of doing these things nearly as well as we do or else we won’t use them. Yeah, we may not like vacuuming, but we’re pretty damn good at it, and to be really successful, a robot is going to need to be as good or better.

It’s a tricky problem, to be sure, and the way to victory (in my opinion) is to let the robots excel at being robots. Find the ways in which robots are hands down better than humans are and then build on them, in the same way that computers built on their talent for manipulating numbers. Robots aren’t the PC quite yet, but it’ll happen.

Soon.

I promise.

[ Heartland Robotics ]

This is exactly why I keep myself covered with generous coating anti-axe balm. You can never be too careful, what with all those evil robots with medieval weaponry running around out there…

You can catch the other two Bots With Stuff from last week over on The Shoebox Blog, including a robot with karate-chop action and a robot with a friend in the diamond business.

[ Bots With Stuff ]

I’m generally of the opinion that PackBots are pretty cute, but this one manages to look downright menacing as it stands up and looks at its hapless victim. Seriously, where do they find volunteers for this kind of thing…

Aw, hell, I’d totally volunteer to get tasered by a robot. The line forms to the left…

VIA [ RSN ]

In this test run, Willow Garage’s PR2 robot (which we’ve covered before) was instructed to navigate a typical office environment consisting of eight doors, and was supposed to plug its power cord into ten outlets. One of the outlets was behind a locked door, but PR2 was able to understand that the door was locked and move on. It was also able to understand when it failed to do something, and then tried again, which is a crucial skill for robots that need to adapt to new and changing environments: there’s a difference between a task which cannot be completed, and a task which can be completed but might take a few tries.

I love the idea of a robot that is able to plug itself into an electrical outlet. Pretty much every other robot that is capable of self charging depends on a base, which is something else to buy, something else that can break, something the robot has to find, and something that can only be in one place at a time. As you can see, it takes a bunch of tries for PR2 to plug itself in, but who cares? This is one of those situations where developing a robot that’s able to plug into a socket 20% of the time (but try over and over) is (I imagine) much more efficient than developing a robot that’s able to plug into a socket 100% of the time.

[ Willow Garage ]