Haptics: Shaking Hands with a Robot
Human beings can be just plain weird on the subject of imitation people. Consider that it was while watching the automata at the palace of Versailles that Descartes decided that the whole canon of Western philosophy needed a rethink.
I always imagine him sitting down to ponder the brainless clockwork mannequins of the French court, trying to find some reliable benchmarks that he could use to tell the metal ones from those that were biological. Or failing that, how he himself differed from the metal figures. Did he stare into some unblinking mechanical eye and say with triumph, "I think, therefore I am"?
Since that time, we've had a slow climb from automata to robots to androids, but a recent survey from the United Nations Economic Commission for Europe (UN/ECE) says we are on the brink of robot boom time. Stable or falling costs for robots, the rising cost of labor, and improvements in technology have resulted in record numbers of robots ordered, according to a Wired News article, "When Robots Rule the World."
Small wonder that other thinkers are taking time out to ponder what all this means. They run the gamut from Ray Kurzweil (whose book The Age of Spiritual Machines draws an optimistic picture of humans embracing a cyborg destiny) to Marshall Brain (whose Robotic Nation site warns that full human membership in a world of robots means that at least half of us will get the idle time we have been hoping foronly we'll get it in the unemployment line).
I recommend the Robotic Nation essay site to those who haven't already looked it over, not because I agree with all he says, but because if ever I saw a site that will make almost anyone want to argue, this is it. But whether you end up grousing about Brain's vision of the future that includes unemployment rates of 50 percent, his proposed social engineering, or his view of the technology, you'll have to agree that he has an important point: Investing some thought up front in how robots will change our society might be a good plan.
According to Marshall Brain, android development pretty well hinges on developing adequate processing power and identifying vision algorithms for interpreting what is artificially "seen." I'm not convinced that they are the only issues holding us back. The world of robotics is dealing with plenty of other challenges. Among them is the need for bots to lift something as heavy as, say, a bowl of cereal, manipulate it the way that humans can (pouring in the milk, taking it to the table), and neither squeeze it till it breaks nor drop it so it shatters. We have a ways to go yet in that direction. And that, of course, brings us to robot hapticsand what they can do right now that is helping to propel us into our brave new robotic world.
In the Hands of our Successors
Robotic hands have long been taking on those industrial chores that are too dangerous, heavy, or monotonous for humans. These days, they frequently also take on tasks that are too delicate for us.
At Johns Hopkins University, researchers are using ever-more-sophisticated systems to do tricky bits of surgery. Current strides in using robotic systems with haptic feedback to conduct laparoscopic surgery fall short when these same techniques are applied to heart surgery because the delicate business of tying off sutures is still at the state where polypropylene stitches break, tearing delicate tissue. That kind of slipped stitch creates serious concerns in heart surgery; this mishap can cause permanent injury or death.
The problem is that the current practice of slicing through the sternum to do heart surgery is invasive. Using the robotic da Vinci system, developed by Intuitive Surgical, surgeons can perform cardiac surgery with less pain and shorter recovery time for patients, despite the concern with sutures. So it was obviously imperative to reduce the risks of pulling and tearing that stitches posed.
To the da Vinci system, doctors work tools at a surgeon's console. Using video overlay optics, the system projects a live video of the operation site right on top of the operator's hands. The larger actions performed by the surgeon are then reflected in scaled-down, microscopic actions by robotic manipulators inside the patient, making very delicate movement possible. But that isn't really good enough for the sutures in cardiac surgery. So researchers David Yuh, Assistant Professor of Cardiac Surgery, and Allison Okamura, Assistant Professor of Mechanical Engineering at Johns Hopkins, have teamed up to work on a haptic interface enhancement. Right now, the experimenters are using a visual force-feedback field to help with the tricky process of knot-tying, but it is accepted as an interim solution until the delicate haptic interface is complete that will allow a surgeon to feel the exact tension of the suture as he or she ties the knot.
And if very tiny surgery isn't strange enough for you, consider the robohaptics going on under the sea. In NASA's NEEMO 7 project, a robotic surgical system called Zeus recently aided a non-surgical team in the removal of a gall bladder. Compared with heart surgery, not altogether amazing, eh? Well, the surgeon (Dr Mehran Anvari) issued his commands from 1300 miles away in Canada. Still not impressed? The "patient" (a surgical training dummy) was in an underwater facility off the coast of Florida! And, by the way, the operation was a success, and reports say that the patient's fine. The exciting implication of this experiment suggests that robot-assisted telesurgery in space might be an option for those who are really far from specialized medical help.
In the realm of haptic hand-holding, there are other subtle technology triumphs afoot. Sommer Gentry of MIT has been busy creating a haptic hand that sends dance signals from one partner to the other. Gentry and her husband, Dorry Segev, a chief resident in surgery at Johns Hopkins, have a remarkable record in swing-dance competitions nationally and internationally, so she knows something about it. So back in 2002, she combined her interests in dance and robots to program a PHANToM robotic device. The hand sent the signals of a lead dancer well enough that human partners holding it could "follow" with the appropriate swing steps.
Since that time, she has been refining her work. She wanted to separate the haptic element from other considerations. Did the musical cues, she wondered, influence the human choice for dance steps?
To test, she provided two human dance partners with different tunes, anticipating some confusion in the ensuing steps. Oddly, leading and following took place much as before, causing her to conclude that touch and a metronome would be enough to communicate dance intent.
"This is encouraging," says Gentry, "because it allows separation of the rhythmic motion cues from the musical meaning of a song. A robot could 'dance' through a task in cooperation with a human without requiring any model for how a human will interpret the emotion or lyrics of the music."
Mind you, the dancing robot feet aren't ready yet, but give this lady a bit more time, and I'm betting that the result will never once step on your toes.