Science fiction has long promised us prosthetic limbs that can meet or exceed the ability of their flesh and blood counterparts. Videogames and movies like Deus Exand I, Robot have squeezed a lot of mileage out of this trope. For many years, though, this idea has been a bit of a sick joke for real amputees, who are often stuck using simple mechanical claws.
Finally, the technology has begun to catch up to the vision of science fiction. This week, DARPA announced a new breakthrough that brings a natural sense of touch to these mechanical limbs.
For legs, the replacements are already pretty sophisticated. Here’s prostheticist Hugh Herr showing off his sophisticated robotic legs, which can allow users to walk, run, jump, and even dance, perfectly naturally.
Unfortunately, arms have proved a tougher challenge. Legs serve a well-defined function: keep the user upright during motion. This goal is so simple that the limb can usually infer what it needs to so from context, using just a few sensors. In contrast, we use arms for all sorts of things, which makes them much more difficult to control.
Early robotic limbs had only a few degrees of freedom — maybe an elbow, a spinning wrist, and a grabbing mechanism. These could be controlled using, say, buttons inside a shoe. This works for some things — holding an object, or even shaking hands — but fails for more unusual or delicate tasks. As robotics has grown more sophisticated, methods of controlling these arms have been forced to improve.
One of the primary innovators in this area has been DARPA, a branch of DoD that researches futuristic technologies, that is developing prosthetics for injured veterans. Here’s one of their most recent prototype arms, which is as strong and nearly as dexterous as a real human limb.
Obviously, this can’t be controlled through buttons in your shoes! The two primary modern control techniques both involve directly interfacing with the nervous system.
Targeted re-enervation works by severing the nerves that drive the missing limb, and routing them into the muscle tissue of the chest. Then, when the user tries to move different joints in the arm, tiny bits of chest muscle flex instead. These flexes can be detected, and used to control the arm. For users who have a substantial amount of arm left, it may also be possible to pick up the flexing of the muscles that remain, without using surgery.
Here’s PBS’ Miles O’Brian trying out the DARPA arm using this technique:
The other option is brain electrodes. For this procedure, the user undergoes surgery to implant an electrode array a few millimeters long onto the area of the brain involved in controlling your hands. The electrode array has hundreds of individual electrodes, each which record different patterns of neurons. A machine learning algorithm can be trained with this data to allow the arm to interpret the will of the user and act accordingly.
Unfortunately, by itself, this isn’t enough to naturally control a limb. Humans rely on their sense of touch and how their limbs are posed (proprioception) in order to perform nearly every task. Being forced to rely just on visual information makes our limbs clumsy and cumbersome.
So how do you add a sense of touch a robotic arm?
The sensors are easy, but getting the information back into the body is challenging. One technology under development is attaching electrodes to the nerve bundles in the remaining limb. Unfortunately, this limits the number of touch channels to just a few — the feedback is relatively crude. Here’s Miles O’Brian again, interviewing an amputee who had these electrodes implanted.
A different approach, and the one that DARPA recently made a breakthrough with, is to go back to the brain, and use electrodes to stimulate it, “closing the loop” so that the arm can be controlled, and its sensors felt, through direct neural interface. This is a lot more natural, and a lot more fine-grained, than what has been possible before.
According to DARPA’s announcement, the user, a 28-year-old quadriplegic man was able to feel and control a robotic limb with great precision, using an electrode array implanted in his brain. The user, who was blindfolded, was able to determine which finger was being touched with nearly perfect accuracy, and reported that it felt like his own organic fingers were being stimulated. It was the first time he’d be able to feel his fingers in more than ten years.
He was even able, when the researchers touched two fingers, to notice that something was wrong, and asked whether they were trying to play a trick on him. This is an indication of just how precise this synthetic sense of touch is. In the not too distant future, more sophisticated brain implants could allow users to perform intricate tasks, like folding origami or playing a musical instrument.
The Path to Market
Unfortunately, there’s a long road from this kind of lab success to actually changing the lives of amputees. While this sort of cutting edge research is exciting, many amputees are still living with simple, mechanical split-hook prostheses. These simple systems were originally developed during the civil war, and haven’t advanced much since. The gulf between what is possible and what people are actually using is huge. It’ll be some years yet before it’s possible to close to gap for most amputees.
What do you think ab