Bionic Technologies Use AI, IoT to Breathe Life into Machines

Imagine a thin robotic finger that can feel and respond to pressure. Or an artificial limb that can lift a can without specific coding. These technologies are being tested as artificial intelligence and deep learning usher in new advancements in robotics.

 

Article Key

Image credit: Dr. Max Ortiz Catalan

Researchers are bridging the gap between man and machine. Prosthetic developers are creating systems that allow patients to control robotic limbs with their thoughts. Initially designed to combat phantom pain, the very real sensation of pain in a missing limb, researchers are adapting the technology to enable greater capabilities.

The new prosthetics combine neuromuscular interfaces and neurotechnology interfaces to create a new kind of neural network. The technology provides the nerve endings with an outlet for brain signals. The approaches vary, but the end result allows for more fluid motion and added dexterity with the prosthetic.

An OPRA Production

Researchers at Integrum AB and the Chalmers University of Technology, both based in Sweden, have developed a prosthetic hand that can be controlled by human thoughts. Using Integrum’s OPRA prosthetic implant procedure, titanium rods are implanted in the patient’s forearm bones. Neuromuscular electrodes are extended to the nerves and muscles in the limb. The electrodes read and react to movements, enabling the patient to control a robotic hand and receive tactile sensations.

Prior to use, the patient practices with a software-based virtual hand. The electrodes in the patient’s limb are connected to a small device that records the muscle activity as the patient views the virtual hand on screen. That tricks the brain into believing a hand exists where it does not.

Man examines a woman's prosthetic arm

Image credit: Nick Beardslee

Mind Control to Major Tom’s Arm

Similar projects are underway at other research institutions. Johns Hopkins is working on a project, sponsored by the U.S. Defense Advanced Research Projects Agency (DARPA), to develop a brain-controlled prosthetic arm for upper limb amputees. The patient wears a modular prosthetic limb that intercepts nerve signals and converts them into motion. Nerves from the missing fingers are embedded near the skin on the upper arm. The patient wears a band of sensors around the limb, which records the patient’s muscle patterns as the prosthetic arm moves. Those patterns can be repeated and trained to move the arm, hand, and fingers without external electrodes.

The hundreds of sensors record temperature, force, contact, and other tactile feedback, which is sent to the brain to create a sense of touch. One patient could pinch two robotic fingers together, hold a small ball, and even feel pressure when a robotic finger was pushed.

In a separate Johns Hopkins project, a double amputee had six electrode arrays implanted into his brain. He was able to control his two prosthetic arms using a brain-machine interface.

Separate but Stronger

Another project funded by DARPA is underway at the University of Michigan. The research team found that by separating the thick nerve bundles into smaller fibers, they enable more precise motor control and amplify the nerve signals. Most nerve signals on amputated limbs are very weak and give off a lot of noise. The research team circumvents this problem by placing tiny muscle grafts around the nerve endings, which amplify the signal and prevent phantom pain from occurring.

Embedded electrodes recorded the grafted nerve signals in the millivolt range, rather than the 5 to 50 microvolt range. That is strong enough to allow the team to record and interpret the signals and then use machine learning algorithms to translate the neural signals into movement. The team plans to begin work on an implanted device, similar to a pacemaker, that would enable thought-controlled prosthetic movement.

“You can make a prosthetic hand do a lot of things, but that doesn’t mean the person is intuitively controlling it,” Cynthia Chestek, Ph.D., associate professor of biomedical engineering at the U-M College of Engineering, has said. The difference with this approach is that the person simply thinks about moving, and the prosthetic limb reacts to it.

The nerve-interfacing technologies allow patients with artificial limbs the ability to perform fluid, coordinated movements using the brain's natural sensory system. They expand the wearer’s functionality and bring life to a robotic body part.

  • Learn more about the research at Integrum AB and at Chalmers University of Technology.
  • Find out more about prosthetic innovations at Johns Hopkins.
  • Discover more about the University of Michigan prosthetic research
  • Hear Stacey Shulman, Vice President of the Internet of Things Group and GM of Health, Life Sciences, and Emerging Technologies at Intel Corporation, talk in our IoT Integrator Wire podcast about technological breakthroughs in healthcare that use AI, including one of her passion projects, Bionic Sight, which gives sight to the blind without a surgical procedure.  

 

Find Intel® IoT Market Ready Solutions

Search the Intel® IoT partner ecosystem and discover 130+ end-to-end IoT solutions.

Become a Member of the Intel® Partner Alliance

If you are delivering repeatable and scalable IoT offerings using Intel technology, the Intel® Partner Alliance program gives you more opportunities to build your business. Get started.