The ultimate goal of bionic hands is to replicate natural abilities. For this to happen, bionic hands need to provide sensory feedback to end-users. This is not only a matter of reconnecting users to their surroundings. It is also critical for perfecting user control.
Sorting Out the Different Types of Feedback
Not all of the feedback available to bionic foot users matches the traditional definition of sensory input. Osseointegration and the Agonist-Antagonist Myoneural Interface (AMI) provide feedback that may be just as important.
However, in this article, we’re going to focus mainly on the traditional forms of feedback so that users who have not undergone osseointegration or AMI may clearly understand their sensory options.
A Daunting Task
The human sense of touch is remarkable. There are more than 3,000 touch receptors on a human fingertip. We can feel not only the size and shape of an object but also its texture, pliability, temperature, and more. Even blindfolded, most of us can identify common objects through touch alone.
We do not just gather this information to satisfy our curiosity. When it comes to using our hands, sensory feedback plays a crucial role in guiding our movements. For example, it helps us manipulate an object into the ideal position for a specific task. This is possible because the brain’s sensory cortex and motor cortex (responsible for movement) share an intimate relationship:
Without these capabilities, users of bionic hands are forced to visually guide every nuance of hand movement — a process that is often slow, frustrating, and tiring.
Unfortunately, it’s going to take a while for bionic hands to provide the same level of sensory feedback as their natural counterparts. First, we’ll need to take some baby steps.
The Use of Pressure Sensors in Commercial Bionic Hands
In September 2019, Psyonic entered the bionic hand market with its Ability Hand.
They took the same sensors found in modern cell phones and repurposed them as highly sensitive touch sensors, which they placed in the Ability Hand’s fingertips.
When the fingers touch an object, they send a signal that triggers vibrations against the skin on the residual limb, informing the user of both contact and pressure.
This makes the use of the Ability Hand more intuitive, as demonstrated in this short video of a user picking up an eggshell while blindfolded:
This doesn’t fully replicate natural touch but it’s an important leap forward in practical functionality. And at $10 for the sensors, no one can argue with its cost-effectiveness.
We believe that the Vincent Evolution offers similar capabilities but have been unable to confirm this due to a lack of English documents and videos.
Others will follow suit or are already doing so.
E-Dermis Electronic Skin
John Hopkins University (JHU) has created an electronic skin called “e-dermis”. This video from freethink.com provides an excellent overview:
This represents a giant technological leap over the use of simple pressure sensors. By embedding sensors in multiple layers, e-dermis has the potential to detect not only pressure but also objects that would typically cause pain. Eventually, it may even be able to sense temperature.
E-dermis offers significant commercial advantages over other methods of sensory feedback. First, it conveys its sensory information by stimulating nerves in the user’s residual limb through the skin. This non-invasive technique, known as transcutaneous electrical nerve stimulation (TENS), is more advanced than simple vibrations yet a lot less expensive and risky than more invasive techniques.
Another advantage is that e-dermis is not tied to a specific prosthetic device. It can be layered on top of any prosthesis, making it a kind of development platform that can be used by scientific teams around the world for many purposes. This serves to pool resources and accelerate advancements.
E-dermis is not the only option in this category. To read about others, see our article on electronic skin.
However, even if this type of electronic skin technology gives bionic hands a near-natural ability to detect different sensations, trying to convey this information via vibration or TENS may not be enough. Fully re-establishing the intimate relationship between the brain’s sensory and motor cortices will likely require true two-way communication with living nerve bundles, otherwise known as a “neural interface“.
The organization leading the charge on this initiative is the Defense Advanced Research Projects Agency (DARPA).
Like What You’re Reading? Follow Us to Stay Informed:
DARPA’s HAPTIX Initiative
While advanced in many ways, both arms lacked a sense of touch. To address this, DARPA launched its Hand Proprioception and Touch Interfaces (HAPTIX) program in 2015. The goal of this program is to enable precision control of a bionic hand and sensory feedback via a neural interface.
One of the program participants is a biomechanical engineering team at the University of Utah. Its objective was to upgrade the LUKE Arm with these new capabilities:
The advantages of this approach are many. First, it helps amputees identify and manipulate objects much faster than they could without sensory feedback. It also seems to reduce phantom pain from the missing limb — a matter of great importance to amputees.
There are, however, drawbacks. One of these is the cost. Implanting electrodes in a residual limb is expensive.
Another drawback is that the stimulation of nerves by electrodes can be imprecise, leading to erroneous or misplaced sensations, such as mistaking the pinky for the forefinger. To make matters worse, the adult brain seems incapable of adjusting its sensory map to compensate.
Portability is another challenge. Until recently, the hand’s sensory feedback had to be channeled through a laptop, translated into the language of the brain, and then communicated to the target nerve bundle via the implanted electrodes. Scientists have come up with a portable device to handle the translation but it is still in field trials.
Finally, there are potential complications with implanting electrodes in the human body beyond just the basic medical risks of surgery such as infection. Scar tissue can form around electrodes or the electrodes can shift position — either of which may interfere with the reception or transmission of signals.
Meanwhile, the commercial version of the Modular Prosthetic Limb (MPL) is due out in 2022. Called the “Atom Touch“, it will feature more than 200 sensors capable of generating four types of sensory data: contact, force, position, and velocity. To perceive this data, users will currently have to undergo Targeted Sensory Reinnervation (TSR) surgery, which will likely have the same risks and drawbacks as other surgeries. However, the company is also working on the ability to convey contact, force, and heat through the skin without surgery.
Integrum’s e-Opra System
This section does involve osseointegration but it’s a special version that includes explicit sensory feedback from an integrated neural interface:
The appeal of this approach is that, since the patient is undergoing major surgery anyway, adding the neural interface does not incur all of the drawbacks and risks of a separate surgery. This changes the cost/benefit analysis of embracing this form of sensory feedback.
Note, the e-Opra Implant System is not yet available commercially.
Sensory feedback is only one aspect of a deeper connection between a bionic limb and its user. To get the whole picture, see Understanding Bionic Touch.
For more information on upper-limb bionic control systems in general, see Bionic Arm & Hand Control Systems.
For a comprehensive description of all current upper-limb technologies, devices, and research, see our complete guide.