In the near future, we’ll begin issuing commands to our home robots by voice. But this will soon evolve beyond the mere one-way issuance of orders. For robots to be truly useful in our homes. it will be imperative that we be able to carry on actual conversations with them. But, just as you have no patience for slow response times on your computer now, researchers have found that you likewise will have no patience from a slow-to-respond robot. Even if that means filling awkward moments of silence with small talk, as your robot is thinking, humans need to receive near-immediate input. Toshiyuki Shiwa and colleagues at the ATR laboratories in Kyoto, Japan, set out to find out just how quickly domestic robots should respond to their owners’ requests by asking 38 students to give orders to a robot. The robot was set to respond in from anywhere from zero to 5 seconds. The students’ patience began to run out after 2 seconds, with 1 second being the optimal response time. However, and this is very interesting, the impatient students were okay with a slower response time if the robot filled in the time with “filler” words, such as “well” or “er.” When the this occurred, people did not notice the delay. In other word’s, make the robots more human-like. Uh, yeah, um, like, I agree. The study was presented at Human-Robot Interaction 2008 in Amsterdam, the Netherlands.
The Incredibly Creepy, Incredibly Cool Modular Robot Snake
Home robots of the future are going to come in every shape and size, as there is no one form that will be suitable for every task. So, what would you think about a robot taking the form of a snake? Think about the applications from investigating your sewers, air ducts and all sorts of tight places. Or what about a security robot guaranteed to scare the heck out of an intruder while he crawls up his leg? The Carnegie Mellon School of Computer Science’s Biorobotics Laboratory is developing modular snake robots whose highly articulated devices can perform a variety of locomotion capabilities that go beyond the capabilities of conventional wheeled and the recently developed legged robots. Watch the video and think of all the possibilities.
Last night I was watching the excellent new series, Terminator: The Sarah Conner Chronicles. The super-sexy robot from the future, Cameron, played by the lovely Summer Glau, was asked by John Conner if she would ever lie to him. She matter-of-factly said yes, she would. He was her why and she responded that she would lie if it helped her mission. Needless to say, John Conner was taken back by that. Someday soon, you too will have robots in your home. While they hopefully will not be a human-killing machine like Cameron, they will none-the-less be your trusted servants or companions. How much would faith will you place in their veracity?
A post today in Medal Blogging brings up this topic for discussion. He cites the work of Dario Floreano et al. on the use of evolutionary algorithms for evolving robots. Apparently his work is shows that some robots actually learn to lie by giving other robots deceptive information about food location. Posts on the subject can be found at Herself’s Artificial Intelligence and Deception Blog. The referred work is described in the a Discover Magazine article called Robots Evolve and Learn How to Lie.
Of course, we needn’t worry. I’d think that by the time robots are smart enough to lie, humans will be smart enough ot have built a personal home lie detector.
For robots to become more lifelike, they are going to need to replace their metallic “skin” and hard bodies with a more pliable and soft membrane. There have been various research projects that are currently in the works (i.e. Lifelike Robotic Skin Advances), and now a new promising project is called the Softbot is now underway at Tufts Biomimetic Devices Laboratory. Their goal is to carry out research into biologically-based technologies that use soft materials not only for skin, but for the body as well. It’s an unusual place to show off a robot, but their prototype, the Softbot, the world’s first soft-bodied robot, is now on display at New York’s Museum of Modern Art (MoMA) until May 12 in an exhibition called “Design and the Elastic Mind.” The robot, or SoftBot, is a bit less than a foot long and looks like a white, silicone caterpillar. Looks like the RoboWorm is one the way, but creative minds can see this technology moving humanoid companions (if you can our drift) closer to reality.
Parents instinctively and naturally teach babies how to talk, but how does that work, exactly? Researchers at at the University of Plymouth will try to find out. Over the next four years robotics experts will work with language development specialists will work with a one meter tall humanoid baby robot named iCub in an attempt to discover exactly how language development works. The team’s findings will hopefully lead to the development of humanoid robots that think, learn and talk. With a bit more technical jargon, Angelo Cangelosi, Professor in Artificial Intelligence said: “The outcome of the research will define the scientific and technological requirements for the design of humanoid robots able to develop complex behavioral, thinking and communication skills through individual and social learning.” Maybe they should just do want a lot of parents unfortunately do these days, just plop iCub down in front of a TV.
There is a debate going on about the ideal form factor of robots. iRobot, of Roomba and Scooba fame, says that robots should not be of humanoid form. “Forget the anthropomorphic features,” says iRobot CEO Colin Angle. But science fiction legend Isaac Asimov disagrees, believing that human form robots are more practical since everything in the world is already built for the human shape.
But do consumers have a preference? New research provides some answers and, interestingly, it is based upon personality types. A team led by Professor Kerstin Dautenhahn at the University of Hertfordshire’s School of Computer Science concludes that people with more extrovert personalities tend to choose more human-looking and human-sounding, humanoid robots, while more introverted people tend to prefer mechanical-looking robots which by, the way, very well could have a head, but a metal head.
“After years of investigating Human Robot Interaction with hundreds of participants, we have looked at proxemics, an area which has not been studied before, and condensed all of this information into an empirical framework,” added Professor Dautenhahn. “Also, rather than producing a robot and then finding an application for it, we have involved people in the development of these People Bots right from the start.”
For robots to become truly helpful in our homes, they need to not only think for themselves, but they need to be able to communicate with their owners more easily. We’ll need to go beyond programming or controlling them through a remote or computer. Beyond voice. Beyond voice? What would that be? Mind control. Your mind to the robot’s mind. It may be closer than you think.
In a breakthrough experiment, researchers at Duke University Medical Center, working with the Computational Brain Project of the Japan Science and Technology Agency (JST), report that the brain activity of a monkey has been used to control the real-time walking patterns of a robot halfway around the world. This work is related to helping those with paralysis regain the ability to walk through robotic braces, but it has broader implications for general mind control of a robot.
The researchers created a real-time transmission of information that allowed the brain activity of the monkey in North Carolina to control the commands of a robot in Japan so that they walked in complete synchronization. The experiment built on earlier work where monkeys’ brain signals controlled the reaching and grasping movements of a robotic arm. The team will begin work in 2009 to develop prototypes of robotic leg braces for potential use with humans.
But, here’s what is the most interesting thing about this experiment, as described by senior study investigator Miguel Nicolelis, M.D., Ph.D., the Anne W. Deane Professor of Neuroscience at Duke: “The most stunning finding is that when we stopped the treadmill and the monkey ceased to move its legs, it was able to sustain the locomotion of the robot for a few minutes – just by thinking – using only the visual feedback of the robot in Japan.” Mind control. Cool.
In the near future, we’ll begin issuing commands to our home robots by voice. But this will soon evolve beyond the mere one-way issuance of orders. For robots to be truly useful in our homes. it will be imperative that we be able to carry on actual conversations with them. But, just as you have no patience for slow response times on your computer now, researchers have found that you likewise will have no patience from a slow-to-respond robot. Even if that means filling awkward moments of silence with small talk, as your robot is thinking, humans need to receive near-immediate input. Toshiyuki Shiwa and colleagues at the ATR laboratories in Kyoto, Japan, set out to find out just how quickly domestic robots should respond to their owners’ requests by asking 38 students to give orders to a robot. The robot was set to respond in from anywhere from zero to 5 seconds. The students’ patience began to run out after 2 seconds, with 1 second being the optimal response time. However, and this is very interesting, the impatient students were okay with a slower response time if the robot filled in the time with “filler” words, such as “well” or “er.” When the this occurred, people did not notice the delay. In other word’s, make the robots more human-like. Uh, yeah, um, like, I agree. The study was presented at Human-Robot Interaction 2008 in Amsterdam, the Netherlands.
by 
Last night I was watching the excellent new series, 
For robots to become more lifelike, they are going to need to replace their metallic “skin” and hard bodies with a more pliable and soft membrane. There have been various research projects that are currently in the works (i.e. 
Parents instinctively and naturally teach babies how to talk, but how does that work, exactly? 
There is a debate going on about the ideal form factor of robots.
Comments