
PR2: Willow Garage's PR2 includes a mobile base, two arms for manipulation, a suite of sensors and two computers, each with eight processing cores, 24 gigabytes of RAM and two terabytes of hard-disk space. The out-of-the-box robot, which costs $400,000, also features an operating system that handles the robot's computation and hardware manipulation functions.
Image: COURTESY OF WILLOW GARAGE
-
The Best Science Writing Online 2012
Showcasing more than fifty of the most provocative, original, and significant online essays from 2011, The Best Science Writing Online 2012 will change the way...
Read More »
What does 2011 hold for the field of robotics? Plenty, if 2010 is any indication. This will not be the year that mobile, artificially intelligent robot nurses assume the responsibility of caring for the world's growing elderly population, but it does promise to be a pivotal time for the development of the underlying technology that will enable safe and reliable automated elder care, not to mention other services that robots are expected to perform in the coming decade.
Thanks to a standardized platform introduced in 2010, roboticists can now collaborate as never before. Last May, Willow Garage, a Menlo Park, Calif., maker of robot hardware and software, released a test version of its personal robot platform. The PR2 includes a mobile base, two arms for manipulation, a suite of sensors and two computers, each with eight processing cores, 24 gigabytes of RAM and two terabytes of hard-disk space. The out-of-the-box robot, which costs $400,000, also features an operating system that handles the robot's computation and hardware manipulation functions.
"There are a lot of innovations in the PR2, but the most significant thing from my perspective is that it is a standardized, well-designed, well-tested platform that has a whole bunch of software that works right out of the box," says Charles Kemp, an assistant biomedical engineering professor at the Georgia Institute of Technology. "You never had that situation before."
Kemp and his team at Georgia Tech's Healthcare Robotics Lab, which he formed in 2007, are focused on creating robots that can safely and effectively help care for senior citizens. The machines would go beyond current efforts to create bots able to follow the elderly around their homes to provide them with Internet access and remind them to take their medicine. For starters, the Healthcare Robotics Lab researchers want their robots to be able to open doors and drawers to retrieve objects such as pill bottles while being guided by a laser pointer, radio signals or touch.
Kemp's lab is one of 16 institutions that experimented with the PR2 during the latter half of 2010. South Korea's Samsung Electronics is using the PR2 to enhance the company's existing robotics research in a country that hopes to put a robot in every home by 2020. The Bosch Research and Technology Center in Palo Alto, Calif., part of electronics and appliances maker The Bosch Group, has begun a two-year project to integrate its advanced sensor technology—including microelectromechanical system (MEMS), accelerometers, gyroscopes, force sensors, and air-pressure sensors—to improve the PR2's performance and reliability. Other beta-testing sites include Stanford University, the Massachusetts Institute of Technology and the French National Center for Scientific Research's (CNRS) Laboratory of Analysis and Architecture of Systems.
Great GATSBII*
Kemp sees the combination of his PR2, named GATSBII, and a free and open-source robot operating system as a way to accelerate his lab's work with the help of a standardized platform and a budding community of roboticists working with the same tools who can now offer more practical advice to one another. "We're actually releasing things that other people can use, and we're using other people's things," Kemp says.
Prior to GATSBII, Kemp and his team used parts from a variety of suppliers to build three different mobile manipulators. The first EL-E, built in 2007 to perform assistive tasks for sufferers of amyotrophic lateral sclerosis (ALS), which impairs physical motor functions. The researchers have since built two more robots: Dusty, which has a lift tray designed to pick objects up off the floor; and Cody, whose two arms and omnidirectional mobile base resemble those of GATSBII. Unlike GATSBII, Cody is the product of many different manufacturers, including Meka Robotics, which supplied the arms, and Segway, which delivered the omnidirectional mobile base. In the cases of EL-E, Dusty and Cody, Kemp and his team designed the robots and then found the parts they needed to actually build them.
*Correction (1/04/11): This was edited after posting. It originally referred to GATSBII as "Gatsby."




See what we're tweeting about




10 Comments
Add CommentThe problem I forsee is that the basic parts are not modal.and thus interchanmgeable.The second problem is that of memory.Why does every robot have to learn everything from scratch Why can't the information be stored in a mainframe then sent via wifi when it's needed.Thus the weight of a large computer would be unnecessary,thus making individual robots much lighter and cheaper to produce.
Reply | Report Abuse | Link to thisWith remote storage of data or software modules you run into issue of communication speeds and also outages. A bot with it's own data and storage that has lost connection but still has power can still function. This is just the whole "cloud" vs local server issue. Each has advantages and disadvantages.
Reply | Report Abuse | Link to thisEveryone talked about modular cars for about a decade but it never happened. Computers used to be quite a bit more modular than they are now because it just isn't cost affective. The same will go with robots. Too much needs to be specialized. A certain amount of modularity is viable but a lot of function simply isn't viable as an add-on to a generic base.
The great thing about this particular situation is that researchers are sharing actual data and collaborating. That really hasn't happened much in robotics as contrasted with cars for example. Car manufacturers actually share a significant amount of research either directly or via patent registration and leasing.
I suggest the following robotical law: No robot shall have humanlike face.
Reply | Report Abuse | Link to thisFor imagine that otherwise in future a man gives on his wife's birthday her a robot with same face as the woman to do the household and later he is called at his office because the house is on fire. The people of the fire brigade tell him that his wife is rescued though still unconcious. Coming home he discovers that it is the robot that was taken to safety while nothing else was spared by the fire.
I am curious to explore the application of these robots, and what their roles would be in the future. I cannot stop imagining the unlimited possibilities and how they can both make our lives easy, as well as complicated.
Reply | Report Abuse | Link to thisI believe the problem here will not be in producing the technology or the application, although a standarized platform will boost the ability of these types of robots to go mainstream, but it lacks in how to sell this to the consumer. I personally do not want a robot to follow me around all day long, I want robots installed in my appliances and other areas of my house to assist in doing things faster, simpler, more efficiently. A washer that i don't have to sort the clothes, it does it for me, a whole house monitor (gas, temperature, humidity ect) that reports to me, things of this sort, no one is going to want to learn/teach/babysit a robot. They'll just want something that works and that is affordable. For that reason i don't see the personal robot that follows you around like Rosie from the Jetsons.
Reply | Report Abuse | Link to thisI’m just glad to see that the PR2 does have great out of the box functions. The fact that there is collaboration within the robotics field is also a great thing. This outlook on technology should be incorporated more in this highly competitive world. With this collaboration I predict that PR2 should have a greatly reduced price from $400,000.
Reply | Report Abuse | Link to thisTo me, most fascinating part of robotics is the use of artificial intelligence to make them capable to serve the world's elderly growing population. To use robotics efficiently to serve senior citizens is not a simple task, but how they would be intelligent to serve the senior citizens? this indicates the power of software engineering and use of artificial intelligence. Dealing with kids and senior citizens are always a challenging tasks. Despite of its social and human aspects, it would be most emerging technology in near future.
Reply | Report Abuse | Link to thisI think a key problem with an aware home is the actual presence of Robots and the effects of what Robot care will have on the elderly care. While I think robot care is an interesting development, I wonder if those being cared for will feel less cared for over time. It is possible that leaving care to a robot might at first seem great, but those experiencing long term care might feel ignored by society and left to the robots. I would expect that human interaction is a much needed and wanted part of current elderly care for social and emotional purposes.
Reply | Report Abuse | Link to thisA standard platform, as well as a readily available wealth of knowledge for the software interface of that platform, is truly what is needed to usher this technology into the sectors of society that need it most. These robots could be used to assist the elderly, as well as the disabled, and could easily be used to replace the current workers in our Assisted Living environments. Not only would they serve the same functions, but they could also cut down on the amount of abuse and mistreatment seen in poorly staffed facilities. Human interaction would be a necessary addendum to the formula, but this is the foundation on changing the way in which assisted living facilities function currently.
Reply | Report Abuse | Link to thisThe point I was making was that unlike a human child who has to be taught everything from scratch.The experience's of robots can be stored in a mainframe and then downloaded thus removing the need for a huge memory and ram to teach each robot what to do.When something new is learned it is uploaded and sent to all other robots via wifi.While each independent robot would have all this information stored on it's hard drive.Thus each robot would continue to function even if the connection is broke.
Reply | Report Abuse | Link to this