Meka Robotics is renowned for their humanoid robot, the M1 mobile manipulator with omnidirectional wheeled base. However, Meka also has a number of joint collaborations with research labs that have resulted in some cool, new robotic technologies. First up, Meka teamed up with Luis Sentis' Human Centered Robotics Lab at UT Austin to design and build a new legged biped, named Hume, for rough-terrain locomotion. As you can see in the video below, Meka and Luis designed a high-speed, high-strength pair of robot legs that will (hopefully) be paired with a set of Meka's famous series-elastic arms (and thus give BDI's PetMan some healthy competition!). Second, Meka teamed up with Mark Cutkosky from Stanford's Biomimetics and Dexterous Manipulation Lab to build upon Standord's tactile sensor technology. They worked with Charlie Kemp's Healthcare Robotics Lab (at Georgia Tech) to develop the sensor into a tactile-sensing skin for Cody that is now being used to explore cutting-edge robot behaviors for tactile manipulation. Check out the videos of the sensor below, and be on the lookout for cool "manipulation in clutter" videos from Charlie's lab in the coming weeks. So... while Meka develops some great products, they can also work with you to make your robot concepts into beautifully-designed robot realities.
Amazon.com just announced that it was acquiring Kiva Systems for $775 Million in cash. Kiva Systems is one of my favorite "companies that employ robots" (rather than "robotics companies" -- there's a big difference). Kiva is all about improving warehouse logistics and fulfillment, and they happen to employ a large swarm of robots (see video below) to accomplish this. Namely, they use robots to autonomously fetch and arrange "pods" so that human pickers can more efficiently fill orders. Their technology is exceedingly relevant in the modern age of large (internet) fulfillment centers. Frankly, they're a perfect match for Amazon. According to a November report, Boston-based Kiva Systems boasted revenues in excess of $100 Million, had 240 employees, was growing 130% year-over-year, and was a prime example of a venture capital-backed robotics outfit (having received $33 Million in total). Honestly, this acquisition makes me sad... I would have liked to see Kiva go public. Plus, I certainly hope Amazon doesn't hoard Kiva away for itself and rob the world of efficient, robotic warehouses...
Animal behavior scientists strive to understand why and how animals do the things they do. Up until recently, scientists had to rely on natural observations -- fortuitous encounters or staged interactions. But advances in biorobots (mechanical robots that mimic live animals) are giving scientists unprecedented control over experimental variables, allowing them to run studies that would be unfeasible (or dangerous, or inhumane) in the course of normal research. They have already proved valuable for testing various animal communication hypotheses: flocking, mate selection, and animal communication. In this article, we examine new work by researchers from UC Davis and San Diego State University that pits a robotic squirrel, RoboSquirrel, against real-life rattlesnakes to study the subtleties of predator-prey interactions
I've been tracking venture capital (VC) funding of robotics companies for the better part of two years. Based on my (limited) data, VC funding in robotics exceeded $160 Million for 2011. This is just a rounding error compared to VC funding of Internet (web-based) companies, which hit a decade-long high of $6.9 Billion in 2011. My hope is that robotics will get more love in the next year(s), but getting VC funding for robotics is a decidedly tough nut to crack. Robotics companies have large capital requirements for robot hardware, few potential acquirers, and almost no "Google-scale" breakout success stories (ie. IPOs). I mean, c'mon... one of the best known robotics companies, iRobot, has a market cap of just $700 Million. This makes robotics a difficult sell to your typical VC firm. My hope is that this list can give others courage to pursue "swing for the fences" type projects along with a source for robotics-friendly VC firms.
I just received word from Berthold Bäuml, a lead scientist in realtime dynamic motion planning at DLR, that they've developed a new humanoid robot named "Agile Justin." Agile Justin is very similar to Rollin' Justin (the ball-catching, Pulp Fiction-dancing robot), except that it has improved dynamic performance. To test the new hardware, DLR researchers have programmed Agile Justin to throw a baseball. Naturally, since Rollin' Justin is able to catch a baseball (see the DLR project page), researchers set up an impromptu game of "catch" between the two robots -- shown in a sneak peek video below. It sounds like this new system is just ramping up and will be used to push the envelope in terms of full-body control: real-time coordination of hands, arms, torso, and mobile base for dynamic tasks. I'm told that technical details should be forthcoming in academic publications later this year along with demonstrations at Automatica 2012.
I'm a huge fan of TED Talks -- high production quality talks given by the world's thought leaders, recorded and distributed for all after the annual TED conference. I've noticed that robotics is a perennial hot topic, so naturally I thought: "I should build a compendium of TED talks about robotics (all on one page)." And so here it is... all 21 of the TED talks that have graced the TED.com frontpage to date. But alas, TED has expanded beyond the confines of the annual (expensive!) conference. There are now dozens of TEDx events (independently organized TED meetups) around the world, and robots are equally popular at these satellite events. So... I guess this will just have to be turned into a series. These videos should get you started. Enjoy!
There are many cool tech toys on the market... But Cubelets make building robots quick and fun. Cubelets are a new robot construction kit from Modular Robotics. Snap these small magnetic blocks together, and without further ado your robot starts to sense, plan, and act. Your robot's behavior depends entirely on how you've assembled the Cubelets; behaviors emerge from the local interactions between Sense, Think, and Action Blocks -- no single “brain” block and no single “program” controls the robot. For example, a Light Sense Block atop a Drive Action Block makes a light-fearing robot. Turn the Drive Action Block around and it’s a light-lover. The KT06 kit, launching next week at CES in Las Vegas, gets you started with six blocks; meanwhile, the KT01 kit includes a full gamut of Sense, Think, and Action Blocks. Cubelets are great for little kids; they can build their first robot in seconds, but big kids (adults) find Cubelets just as much fun too. This Cubelets video (below) shows how it works.
iRobot has received ample attention for their particle jamming innovations (ie, the "Jamming Blob Robot" and "Jamming Gripper"), created under the now-expired DARPA Chembot program. However, if you're like me, their particle jamming actuators and hexapod "JamBot" probably alluded your attention -- and they're stinkin' cool! That said, I'd like to introduce you to the "Hexapod JamBot" and the "Jamming Modulated Unimorph (JMU) actuator" created by researchers at iRobot and the Jaeger Group at the University of Chicago.
I'm really excited about inflatable robots... they have the potential to be low-cost, lightweight, extremely powerful, and yet "human safe" -- ie. perfect for many robotics applications. With that in mind, I would like to introduce you to two new (breakout) inflatable robots: a 15-foot-long walking robot (a Pneubot named Ant-Roach) and a complete, inflatable robot arm (plus hand). Both of these robots were developed by Otherlab as part of their "pneubotics" project (in collaboration with Meka Robotics and Manu Prakash at Stanford University), with some funding from DARPA's Maximum Mobility and Manipulation (M3) program. These robots use textile-based, inflatable actuators that contract upon inflation into specially-designed shapes to effect motion. Since these robots are built out of lightweight fabric-and-air structural members and powered via pneumatics or hydraulics, they exhibit large strength-to-weight ratios. For example, Ant-Roach is less than 70 lbs and can probably support up to 1000 lbs; the inflatable robot arm is less than 2 lbs and can lift a few hundred pounds at 50-60 psi. Be sure to read on for details and lots of videos!
This article is specifically for folks in academia... When writing a journal paper, targeting the right venue is an important consideration. There are lots of factors that go into this decision: audience, prestige, historical topics of interest, turn-around time, open access, etc. Discussing all the considerations in detail is too taxing and is probably not actionable (it's too dependent on your research and goals). But I thought I'd share... I'm tracking the Institute for Scientific Information (ISI) impact factors for various robotics journals. In very general terms, the impact factors can give you a rough approximation of journal quality and help you target your publications. You can find a historical plot of robotics journals' impact factors (along with the latest values) below. I'll try to keep these up to date.
I would like to introduce you to a new "elastomeric rolling robot" -- a soft robot made of inflatable, silicone actuators that pressurize in sequence to make the robot move. This new robot hails from MIT's Distributed Robotics Laboratory and has a major distinguishing feature compared to other soft robots: it is entirely self-contained -- no more off-board electronics or pneumatics; everything is on-board. Two technologies facilitated this new robot: (1) A "pneumatic battery" that uses mechanical feedback to self-regulate a chemical (hydrogen peroxide) reaction and maintain a stable pressure inside the robot's on-board pressure vessel. (2) An energy-efficient pneumatic valve design based on electropermanent magnets (one of my favorite topics!). These two new technologies were just presented at recent robotics conferences (ISRR 2011 and IROS 2011). Be sure to check out the video below.
Vibrobots (and bristlebots) are simple robots that use a tiny pager / cellphone vibrator motor (with an eccentric weight) to randomly bounce around -- they are the subject of many Maker / DIY projects as well as some well-known commercial toys (such as the $7.00 Hexbug Nano). Naghi Sotoudeh, a Hizook reader from Iran, contacted us about his latest project: a remote controlled micro-scale vibrobot (measuring just 18 x 12 x 10 mm) that uses two vibrator motors to achieve steerable motion without any wheels. Naghi's design is similar to some previous steerable vibrobots (eg. the Harvard Kilobot project), but the vibrator motor arrangement gives his design a nice, distinctive faux-wheel look. The hardware is fairly simple: a small PCB, two vibrator motors, a microcontroller, an infrared photodiode, and a very small battery. In general, the software for this type of robot isn't too bad either. In short, this could be a great DIY project and potentially a nice mass-market product. What do you think... Would you fork over your hard-earned money for a RC vibrobot kit or pre-built RC vibrobot?
It's that time of year again... MIT Technology Review announced their 2011 "Young Innovators Under 35" Awards (TR35). This year two roboticists are among the recipients: Brian Gerkey and Pieter Abbeel. Brian Gerkey is currently the "Director of Open Source Development" at Willow Garage, where he architects ROS (the Robot Operating System). ROS is quickly becoming the world's standard robot software platform, supplanting Player --which was also developed by Brian. Pieter Abbeel, a professor at UC Berkeley, has done some cool stuff with the PR2 (eg. towel folding) as well as really nice machine learning work on autonomous helicopter acrobatics. Now we can add Brian and Pieter to the ranks of past TR35 robotics recipients: Aaron Dollar (2010), Andrea Thomaz (2009), Andrew Ng (2008), Robert Wood (2008), Josh Bongard (2007). I'm noticing a nice trend... Hopefully TR35's love for robotics continues.
The Swarmanoid project is a cool twist on swarm robotics -- researchers use a heterogeneous swarm of robots to achieve distributed mobile manipulation. The swarm is comprised of three different robot varieties: Hand-Bots (manipulation and climbing), Foot-Bots (wheeled mobility and sensing), and Eye-Bots (quadrotors for recon and sensing). The latest video of Swarmanoid retrieving a book won the "Best Video Award" at the Artificial Intelligence Conference (AAAI 2011) in San Francisco just the other day. You can check out the robots and winning video below.
News just broke on XinHuaNet (via Slashdot) that Foxconn, a Taiwanese company with more than 1M Chinese laborers on the mainland, plans to deploy one million robots(!) over the next three years -- a 100-fold increase over current numbers. This should serve as a wake-up call for the United States. China already dominates in manufacturing; if they can capture the "new" flexible, light manufacturing space too, then the United States will be in dire straits (National Robotics Initiative or not). One commentor on HackerNews suggests that the robots will be ABB's Frida. Of course this needs more substantiation, but ABB isn't exactly a newcomer to industrial robotics; the Swiss company has been around for ages. Still, it would be mildly surprising if ABB wins out over all the competition (eg. Heartland Robotics) that are specifically trying to establish themselves as pioneers in "flexible, light manufacturing." The future of robotics certainly looks interesting!
Robotis has just announced a new line of Dynamixel Servos: the MX-series, beginning with the release of the MX-28. The MX-28 servo has some distinct improvements over its RX-28 brethren: 12-bit (0.088°) angular resolution (four times that of the RX), full 360° controlled rotation (rather than 300°), non-contact magnetic encoders (not subject to mechanical wear), etc. -- and all for about the same price (MX-28: $219.90 MSRP, RX-28: $200)! Of note, the MX-28 is prominently featured as part of the new DARwin-OP humanoid -- the recent Nao competitor that Robotis created in collaboration with Virginia Tech's RoMeLa Lab. Perhaps best of all... Hizook was selected as a beta tester for this new servo (probably owing to our prominent coverage of the RX-series and our awesome cross-platform open-source Robotis software library). We were impressed with the new MX-28 -- read on for details, including an exclusive look inside the new servo as well as a quick tutorial using the updated open-source drivers (in python, complete with ROS bindings).
ICRA 2011 kicks off in less than two weeks in Shanghai, China. Unfortunately, I won't be attending -- I'll have to wait to see y'all at IROS 2011 in San Francisco come September. However, I was browsing through the program and it looks like there are going to be a lot of great talks! A few that caught my eye: Morgan Quigley's low-cost 7 degree-of-freedom (7-DoF) series elastic actuator (SEA) arm for sub-$5k, Aaron Dollar's helicopter-mounted shape deposition manufacturing (SDM) hand, and Rollin' Justin's fast reflexes to catch flying (tossed) balls... and these are just the talks that have videos posted online already (embedded below). Eventually I'll get around to writing about all this cool research (after finishing my dissertation in a few months!) -- but for the time being, here's a quick scatter-shot overview.
Carol Reiley, a surgical robotics PhD student at Johns Hopkins' Computational Interaction and Robotics Laboratory, wrote in to share how they unwind after a long day of research. She writes, "Here's a video of the Johns Hopkins Robotics lab playing the board game Operation on the $1.3M dollar (Da Vinci Surgical) robot. The video emphasizes the robot's precision and hand tremor reduction as well as how difficult Operation is to play, even with a robot. :-)" Fun, whimsical, and a bit meta. Surgical roboticists seem to enjoy themselves; seems like just yesterday they were using the robot to make little paper airplanes or produce dance videos.
Perhaps you remember Garratt Gallagher -- he's the ROS / Kinect hacker responsible for 30% of all entries in the recent Kinect / ROS 3D competition, in which he won first place for Impromptu Buttons (his other entries such as Finger Detection, People Follower, Kinect Minority Report Interface, etc. were also quite impressive). However I want to introduce you to his most exciting project to date: Bilibot. Started in December and funded with $5,500 through KickStarter, the Bilibot Project is an open source effort to create an affordable robotics platform based on the iRobot Create, Kinect, and a computer pre-installed and pre-configured with Ubuntu and ROS. The goal is to create a platform for hackers, enthusiasts, and researchers that works right out of the box without the (often daunting) challenge of installing and configuring Ubuntu and ROS. Frankly, the $150 Kinect may be popular in the gaming industry, but it will completely revolutionize robotics -- it is an amazing sensor, and the Bilibot project aims to make it even more accessible.
This is a guest essay by Fred Nikgohar. Fred is the CEO of RoboDynamics, makers of the TiLR robot which was recently featured in a New York Times overview of telepresence robots. Fred argues that we've reached a watershed moment in robotics facilitated by cheap 3D sensors like Microsoft's Kinect (ie. PrimeSense RGB-D camera) -- that the Kinect provides a roadmap where "the best solution to complex, low-cost sensing (or actuation for that matter) is to take advantage of affordable, mass-produced components, complementing them with the innovative use of software solutions that benefit from constantly declining prices of computation."
I would like to introduce you to iRobot's latest prototype: a new telepresence robot named AVA that was unveiled this week at the Consumer Electronics Show (CES 2011) -- see the video below. Through AVA, iRobot intends to explicitly leverage the proliferation of tablets / smart phones and their associated app stores. They intend to furnish actuation (mobile base, pan-tilt unit, telescoping linear actuator spine, etc.), a sensor suite (including sonar, laser, and a depth camera like Microsoft's Kinect), and basic robot software (eg. obstacle avoidance, mapping, and direct physical interfaces). Meanwhile, you provide the brains in the form of a tablet (eg. iPad or Android). In theory, this should open up mobile robot application development to a much broader audience, creating the oft-discussed robot app store. When combined with the recent announcement of the Scooba 230 floor cleaning robot (which I will certainly purchase), I would say iRobot is still innovating!
Many folks visiting Hizook today are looking for a "robot coffee machines," specifically, the Tassimo BrewBot by Bosch. In actuality, BrewBot is not a robot at all! They are using a cute robot coffee machine to sell decidedly non-robot coffee makers, and it makes me sad... I want the actual robot! Either way, kudos to the clever marketers at Bosch for making such a great commercial (embedded below). In the meantime, if you're interested in seeing a real coffee making and delivering robot, go check out the Nestlé Nespresso Nesbot. Dang it, now I've got a hankering for an espresso...
Take a moment and envision an electromagnet: a simple coiled wire driven by a hefty electrical current gives a fully-programmable magnetic field strength (on, off, and everything between). Electromagnets are ubiquitous, but it turns out that there is a little-known device with similar functionality yet zero static power consumption -- they are called electropermanent magnets, and they've been around and in use since the 1960's! A 2010 PhD thesis by MIT Media Lab's Ara Knaian examines the physics, scaling, trade-offs, and several new actuator designs (eg. stepper motors) using these little-known wonders. Recently, electropermanent magnets facilitated an innovation in "programmable matter," where they were instrumental in creating the world's smallest self-contained modular robots to date (12mm/side). Read on for details about this fascinating technology, along with discussions about existing and possible robotic applications.
Heartland Robotics, the stealthy robotics startup founded by iRobot co-founder and robotics legend Rod Brooks, was in the news again last week after closing a $20M financing round. Little is known about the company beyond broad superlatives from executives about building robots to "increase productivity and revitalize manufacturing." Now, successful fundraising by a robotics startup is great news, but alone it was insufficient to draw my laser-focus away from thesis work. However, a Boston.com article this weekend provided a tantalizing new nugget of information that I absolutely must share -- Heartland is working on a new mobile manipulator with a $5,000 projected price point complete with one or two arms, grippers, sensor head, and a mobile base. If coupled with a depth camera (eg. Kinect) and a decent computer, this could be a really compelling robot platform! If this price point is real, perhaps those superlatives aren't so inflated after all...