Rethink Robotics is giving researchers an ability to get up close and personal with its Baxter industrial robot, with a new research version they can use to build custom software and applications for the robot.
The Baxter Research Robot -- a two-armed humanoid robot much like its commercial counterpart and with the same $22,000 price tag -- includes a software development kit designed for ROS, the Robot Operating System standard on which Baxter runs. Developers can use the software to develop custom applications that leverage the robot’s built-in sensors, such as cameras, force, sonar, and rangefinder.
Rethink Robotics is offering a research version of its Baxter industrial robot powered by a software development kit. The robot costs $22,000 and is aimed at giving robotics researchers an opportunity to create new software for Baxter, a robot designed to work side-by-side with humans safety and intuitively. (Source: Rethink Robotics)
“Our goal with launching the Baxter Research Robot is to offer an affordable system for researchers to develop apps to perform a myriad of tasks,” Mitch Rosenberg, Rethink's vice president of marketing and product management, told Design News in an email:
Of course, many researchers will be manufacturing engineering teams at large manufacturers. We’re ultimately hoping our introduction of an open hardware platform will result in big ideas in terms of software applications and end effectors, much like software did for computers and apps for smartphones. The possibilities are limitless.
The hardware for the research version of Baxter is exactly the same on both versions of the robot; the only difference is in the software that is installed, Rosenberg told us. While the manufacturing version of Baxter is pre-programmed for training in industrial work, the research version comes with the SDK, giving developers a blank canvas to program the robot. Developers can communicate with Baxter’s development workstation via an Ethernet connection and a network address, Rosenberg said.
Though he wouldn’t venture to guess what kind of applications might come from offering a programmable Baxter to the research community, Rosenberg said there are a number of aspects of the robot’s functionality that can be developed using the SDK. “Researchers are able to focus on fundamental goals such as object manipulation, human-robot interactions, collaborative robotics, robot adaptation to changes in its environment, and arm trajectory planning.”
Researchers at MIT and Tufts, among others, are already working with the Baxter Research Robot.
Rethink Robotics unveiled Baxter last year and positioned the robot as a major step forward for industrial robot design. Key to Baxter’s design is its more humanoid appearance and ability to work safely alongside humans on the factory floor by sensing its environment through built-in software and sensors. The robot also is easy to train, so it can take on an assortment of tasks. Traditionally, industrial robots have looked more machine-like, often have to be caged off to protect human workers from any possible malfunction, and are programmed to perform one specific function.
This is a great idea. Open software for the robot could create possibilities far beyond what Baxter would have developed for its machines. This open model worked very well for Google when it gave away its Android smartphone operating system. The results were robust enough to allow Google to go toe-to-toe with Apple -- no easy feat. Because of its open approach, Android is now the leading smartphone OS in volume.
Maybe this post should be titled "Rethink Rethinks Baxter." Anyway, thanks for the post, Elizabeth. Making Baxter available to developers is a great move on Rethink's part and will help accelerate development of this important robot design platform.
Yes, I agree with all of you, this is a really great idea and it will be interesting to see what developers come up with when they have a chance to program for the actual robot and not just a facsimilie of the platform. Is this a novel idea, does anyone know? Have other companies done similar things? I wasn't sure about that myself.
Elizabeth, I'm not sure what question you're asking. Opening up robot platform development is not a novel idea--that's why ROS exists, and at least one other robot company Willow Garage, has done more or less the same thing, including providing its robot hardware for tinkering. But WG wasn't originally formed to just make robots commercially, so that part is a relatively new idea. Also, ROS has been extended to industrial robots. We've covered the development of open-source robotic development in several posts: http://www.designnews.com/author.asp?section_id=1386&doc_id=245355 http://www.designnews.com/author.asp?
Thanks for the clarification, Ann. What I was asking was somewhere between what both you and Rob answered--ie, has a commercial robot platform been opened up this way. I'm familiar with Willow Garage, and of course as you not there are a lot of open robotics platforms in the hobbyist world mainly, is that right? Sounds like Rethink is doing something a bit different for the commercial platform, which personally I think is a good thing.
That's what I thought you were asking. As a commercial robot supplier--although a very new one--Rethink is definitely doing something different by opening up their hardware platform. This has already been done by researchers designing surgical robots, as we covered here http://www.designnews.com/document.asp?doc_id=239419
Thanks again for the clarification, Ann. Surgical robots seems like a good place to open up the technology as well for the reasons you point out in your story--price point and complexity. Have there been results in terms of how it has advanced that particular platform?
Rob, I asked a question in my last comment and you kind of answer it. Google is one company that used this model to its success. I was wondering if it's ever been done with a robotic platform before. I imagine probably not, but perhaps this will set a precedent for the future, if it goes well. Savvy move, too, on Rethink's part, as they are pushing Baxter as a real innovation and are going for innovation in development as well by offering a research version.
Elizabeth, I'm not aware this open system approach has been used in robotics. I don't think it's been used much at all. I'd like to hear from those who have seen in used in other contexts. Google results were stellar. They gave away the platform and it because the leading platform for smartphones.
This certainly creates a very potent development environment for software developers. Will save alot of integration time developing hardware and testing, especially since there seems to be so much interest in humanoid type robots.
I agree, Apresher, this does make for a potent development environment. It's good to see open source moving into new territory. It will be interesting to see what comes of this. It will depend on whether robotic engineers eagerly take to this model. The jury's out on that question.
I think what makes this unique is that Rethink has developed its system using open software but now is making its humanoid hardware platform (with all of the development that went into it) available for research purposes. Researchers can concentrate solely on software solutions with than many problems already solved for them. Very interesting development.
Yes, that definitely seems like a unique value proposition, Al, especially since hardware and software development are two very different sides of the same coin. Those developing software and those developing hardware are usually not the same folks, so this definitely makes it easier for those on the software side of the house to program Baxter.
Are they robots or androids? We're not exactly sure. Each talking, gesturing Geminoid looks exactly like a real individual, starting with their creator, professor Hiroshi Ishiguro of Osaka University in Japan.
Truchard will be presented the award at the 2014 Golden Mousetrap Awards ceremony during the co-located events Pacific Design & Manufacturing, MD&M West, WestPack, PLASTEC West, Electronics West, ATX West, and AeroCon.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.