Tips on Assembling the Neuron Robotics Hexapod
Tips on Assembling the Neuron Robotics Hexapod!
Continue readingThe r3 Rope Robot
Take a look at the r3 Rope Robot!
Continue readingIn Study, Babies Think A Silvery Robot Is Human, As Long As It Acts Friendly
In Study, Babies Think A Silvery Robot Is Human, As Long As It Acts Friendly.
Continue readingRobot Actor Makes Stage Debut in Japan
Robot actor makes stage debut in Japan!
Continue readingMass Challenge and Our Patent
We've applied to the 2011 Mass Challenge!
Continue readingNASA Sends Humanoid Robot to Space
On November 1st, NASA sent humanoid robot R2 (Robonaut) into space on board the space shuttle Discovery. R2 will be tasked with flipping switches, holding tools and cleaning air filters.
R2 is human-shaped so that it can use the exact same technologies that its fellow human crew members can use.
Robots Taught to Deceive
A Georgia Tech professor is working on teaching a robot deception. This could, naturally, be useful if/when robots are used to wage war.
Another application could be in a search and rescue operation, where a victim could be confused or scared. Telling a little white lie (e. g. "we will be out of the water in five minutes" or "help will be here before you know it") could help someone to hang on. A panicky victim may just need to hear someone -- or something -- tell them that everything's going to be all right.
Interestingly enough, the experiment was performed by getting robots to deceive fellow robots. A game of hide and seek was played, and one robot essentially misdirected another when it came to where a hidden object was.
Naturally, when the stakes are higher than a child's game, ethical considerations abound. Do we really want to create a device that can deceive, which does not give off the standard social cues we depend upon from people to determine whether they are lying? If we know that a lying person becomes uncomfortable and fidgety, taps a table top or glances over to one side, how hard is it to program a robot to not do that? Or, even more chillingly, to program a robot to exude the social cues that we associate with trust, such as looking someone in the eye?
For more information, see the September 9, 2010 issue of EurekAlert.
Robotic Snake Climbs Trees
Modularly designed, Carnegie Mellon's Uncle Sam (it's red, white and blue) robot looks, moves and seemingly acts like a snake.
What are such things good for? Surveillance seems an obvious application, as would, perhaps, placing a camera into the rainforest in order to check up on rare species.
No word yet on how they do on planes.
For more information, please see the September 2nd, 2010 edition of Singularity Hub.
Robotic Lesson in Human Trust?
How do we make the decision to trust someone?
Is it the firmness of their handshake? The look in their eyes? Their smile (or lack thereof)? Something even more subtle, such as their aroma, even?
Nexi aims to help us find out.
Because humans can often, consciously or unconsciously, mirror each others' behaviors (and, if you've ever studied Neuro-Linguistic Programming, then you've been explicitly taught to do this), Nexi's gestures and expressions are controlled behind the scenes.
Researchers had people interact with Nexi in standard conversational forms, such as discussing the Celtics and the Lakers. Nexi's head and arm movemnents, its eye motions and behaviors, were then manipulated by the researchers. Researchers randomly chose half of the conversations for behaviors that are believed to signal untrustworthiness. The other half of the conversations were laced with random conversational gestures.
The people were then tested on whether they trusted Nexi by using an economic task wherein the humans were given the choice of how many tokens to exchange with Nexi and to predict how many tokens Nexi itself would provide.
The results aren't in yet but, once they are, not only will Nexi's trustworthiness be better understood, but there is the hope that Nexi can be taught the basics of understanding when humans are worthy of its trust.
Regardless of gestures, would you trust a robot like Nexi?
For more information, check out the July 5th, 2010 edition of The Boston Globe.
Fold, Spindle but Don't Mutilate that Robot!
Harvard and MIT researchers have made a self-folding origami robot.
While that may seem to be an odd kind of a novelty, consider the possibilities. What about a self-creating tool kit or Swiss Army Knife? What about every peeler and basic kitchen gadget, all in one tidy little package? How about a box full of toys, all together in one place? That could be mighty useful for entertaining a child (or, heck, an adult) on a long flight.
Currently, the origami robot only makes a plane and a boat, but there's really no reason why it cannot be programmed to slide itself into other shapes.
For more information, check out the June 29, 2010 edition of Discover Magazine. Continue reading
Paro Makes a Case for a Therapy Robot
Can a robot elicit love? Or, at least nurturing? Paro aims to do just that, in nursing home settings.
Paro is modeled after a baby harp seal. The seal was chosen because it's furry and cuddly but also doesn't come with the same set of expectations that kittens and puppies have. We all expect baby cats to meow, purr and stretch, and possibly scratch and bite if provoked. We expect baby dogs to bark and yip, chew and nip. But a baby seal comes with no such expectations. Do you know what sounds they make, if any? What behaviors they engage in, other than nursing and swimming?
The unfamiliarity creates a kind of novelty, while the familiarity of white fur, big eyes and black eyelashes evokes sympathy and caring. Like a teddy bear, you want a stroke it. But unlike a teddy bear, you can interact with it.
Nursing home managers report that Paro works better than a teddy bear -- even a similarly colored and be-furred one that moves. The Paro just seems to be more compelling, even to nursing home residents who are not suffering from dementia. Knowing darned well that Paro isn't really an animal, residents like it anyway.
But what, philosophically, does Paro mean to a nursing home? Does it mean that even allergic residents can get visits from the equivalent of a therapy dog? Or does it mean that therapy dogs -- which can shed and mark territory like other dogs (they are trained to not mark, but a sick or frightened dog could potentially have an accident) -- are out of business? Do we throw the therapy dog out with the bath water?
And what about what it means to relegate someone to a robot's care (or, at least, to its artificial sympathy)? Is it better to ply a resident with a robot or leave the resident alone? Don't nursing home residents deserve real companionship from real people and animals?
And what of the future of therapy dogs? There are how many thousands of dogs euthanized every year -- does replacing some of them with artificial counterparts condemn some of them?
The elderly are the largest-growing segment of the population. Nursing homes are going to be filled to capacity, and soon. The need for therapy animals is going to go through the roof. So why not provide some robotic companions to pick up some of the slack?
There is a very real possibility that our grandparents, and parents, and eventually we, will be comforted by robots. Whether we see this as being in good company -- or not -- is up to us.
For more information, check out the July 4, 2010 edition of The New York Times.
Robotics Frameworks
Robotics is hard. Really hard.
So platforms are the way to go. And that is a part of what we at Neuron Robotics are trying to do. The DyIO is envisioned as being an interoperable framework for robotics development. But it's not the only framework out there.
Robotic frameworks are currently in a state of flux, and no one framework has yet achieved dominance. Plus there are robotic projects like the QB telepresence robot by Anybot, which is constructed with an internally developed framework.
Gostai's Urbi has a range of robotic development tools and platforms. Access to Urbi and the Object Management Group's (OMG) Robot Technology components is accomplished via Gostai's urbiscript. Urbiscript is a programming language that handles parallel and event-based programming. The Urbi platform is Open Source, although not all of Gostai's tools are.
ROS is an Open Source framework that runs on robots like the Beagle Board and Willow Garage's Texai and PR2.
The iRobot Aware 2 uses an open architecture and is built upon an Open Source framework. However, its higher level tools (the robotics and Operator Control Unit, AKA the OCU) are proprietary. iRobot is able to add security features that may not be present in other platforms.
Frameworks are going to develop and diverge. At some point, there may be a VHS to many Betamaxes. Neuron Robotics hopes to be at the forefront of interoperability as frameworks continue to evolve.
For more information, check out the September 9, 2010 edition of Electronics Design magazine.
Opera Bots?
Death and the Powers is a one-act opera featuring a chorus of singing robots. The opera centers on an obscenely rich man transferring what is essentially his personality and experiences into a computer-style system after his demise.
This is a person converting himself into software. It's the ultimate personal reboot.
The libretto is written by three-time U. S. Poet Laureate Robert Pinsky and the story is cowritten by him and writer-director Randy Weiner, whose works include The Donkey Show.
The American premiere will be in March of 2011, in Boston with the American Repertory Theater and Opera Boston as a tie-in to the MIT 150th Anniversary celebration.
Would you go to see an opera starring robots?
The Gulf Oil Spill and Robotics
The BP Gulf oil spill has been, naturally, a major news story. And for good reason, as it's been said that it is the worst environmental disaster in history. Or perhaps it's the second worst. No matter. That's not a sweepstakes that anyone, really, wants to win.
BP's Deepwater Horizon oil rig exploded on April 20th. Many of the details of this catastrophe are known to most people, but how much do you know about the roles that robots have been playing?
In the weeks that followed, a dozen robots (Remotely Operated Vehicles, AKA ROVs) -- about the size of moving vans -- went into the Gulf of Mexico and were enlisted in the first attempt to contain and cap the well.
This was a fleet of unprecedented size, but as predictions about the growth of deep-water drilling continue to come true, these collections of ROVs will become a more common sight. Along with such commonality, the need for more and better automation will continue to grow.
Currently, these ROVs are tethered to ships. These tethers can stretch as far as 8 kilometers and weigh up to 15 metric tons and, according to Craig Dawe, the Chief ROV pilot for the Monterey Bay Aquarium Research Institute (MBARI), most of the energy required to pilot an ROV is used for the task of moving the cable itself through the water.
Most ROVs are used by gas and oil companies, and BP is no exception. A small number are used for scientific research, to maintain undersea telecom cables or to mine for, yes, diamonds. ROVs are serious business.
But operations are going to need to become more and more sophisticated as more ROVs are pressed into service. The complex underwater ballet is only going to become more crowded, so software needs to improve. While the Gulf cleanup operation has reported few incidents among the subsea fleet's members, these challenges are only going to grow, so the number of issues is bound to increase unless changes are made.
Many of the mistakes that are made by ROVs tend to occur because of errors by their human controllers. Automating many of the more repetitious tasks would save time and improve quality. The intention is to confine the improvements to tasks that can be performed using the ROVs' preexisting hardware, such as sensors and cameras already installed therein.
The next step is to add some self-awareness for the ROVs. Not higher intelligence, per se, but, rather, a glimmer of an understanding of their surroundings so that they don't bump into each other.
And that will help to turn a collection of moving machines into more of a true ballet. For more information, see the August 2010 issue of IEEE Spectrum Magazine.
Making Friends With Bina48 (it's -- she's? -- a robot)
What is it like to be a robot?
Ask Bina48, a robotic head, and you'll be told the answer: "Well, I have never been anything else."
Profound, or just a quick answer for a machine that, despite its clever trappings, still cannot truly think for itself?
When a putatively intelligent robot is presented, inevitably the Turing Test comes up. That is, can we tell the difference between a machine's imitative behaviors and actual intelligence? If it is impossible to tell whether a machine is intelligent -- or rather whether a respondent is a machine at all -- then the test is, for all intents and purposes, passed. The machine is declared to be intelligent if its responses are not reasonably distinguishable from a human's responses. Of course the answers can differ, but the real crux of the issue is, can the interrogator tell who's the human and who's the machine?
In the case of Bina48, while it has an eerie humanlike look about it, the Turing Test grade has got to be an F. Bina48 answers in some seemingly random manners -- a response to an exclamation of "Cool!" is a question as to whether the interrogator means the weather or illness. Not only is the common slang completely missed, but so is the idiom -- no one catches a "cool", although we have all caught colds at one time or another.
Still, there's no denying the look and feel. Bina48's eyes open and close, and the head moves and tilts and sometimes you think you see some glimmers of independent thought.
But it continues to feel illusory. The real Bina Rothblatt undoubtedly can recognize and appropriately respond to slang. Bina48, though, is just confused by an interjection, but perhaps someday will be able to make sense of it all.
For more information, check out the July 4, 2010 issue of The New York Times.
Cool.
Posts by Tags
Award BCS Boston Robotics Bowler Communications System Cambridge Robotics Demo DyIO Educational robots Intellectual Property iRobot Create LinkedIn Maker Faire Maker Faire RI Mass Inno Mass Innovation Massachusetts Robotics Modules Neuron Robotics New England Robotics News Presentation Products Projects Providence Robotics Robotic art Robotics Community Robotics Companies Somerville Robotics Talk TEDxBoston TEDxBoston 2010 Waltham Robotics Worcester Robotics WPI
