Researchers at the University of Utah are using artificial intelligence (AI) to make bionic hands more intuitive, giving them something closer to a mind of their own.
Engineering professor Jacob A. George and postdoctoral researcher Marshall Trout of the Utah NeuroRobotics Lab integrated AI into a bionic hand fitted with proximity and pressure sensors, giving prosthetic users better grip precision, improved dexterity, and reduced mental effort.
“Nearly half of all prosthetic users abandon their devices, often citing poor controls and cognitive burden,” Trout said. “Our goal was to make the prosthesis feel more like an extension of the user rather than a tool they must constantly command.”
One of the ongoing challenges in prosthetic design is replicating a person’s natural sense of touch. Humans rely on reflexes to grasp objects securely, without consciously thinking through every micro-adjustment. To address this, the Utah team used a TASKA Prosthetics hand with custom fingertips that included optical proximity sensors and pressure detectors. These sensors are sensitive enough to detect even a cotton ball landing on the fingers, giving the AI rich data to predict and adjust each finger’s movement.
In the next phase, the team trained an AI-driven neural network on the sensor data. Each finger operates in parallel, automatically positioning itself to keep objects stable and secure. That raised an important practical question: what happens when the user wants to let go rather than hold on? To solve this, the researchers implemented a shared-control system that blends human and AI inputs, allowing the prosthesis to assist without overriding the user’s intent.
“The study team is also exploring implanted neural interfaces that allow individuals to control prostheses with their mind and even get a sense of touch coming back from this,” George said, adding that next, “the team plans to blend these technologies, so that their enhanced sensors can improve tactile function and the intelligent prosthesis can blend seamlessly with thought-based control.”
The Utah NeuroRobotics Lab tested the system with four transradial amputees, including individuals with amputations between the elbow and wrist. Participants completed dexterity tasks and everyday activities, such as drinking from a cup or picking up small objects. The AI-assisted prosthesis allowed them to perform these tasks easily and reliably, with little to no training.
“By offloading the fine-tuned aspects of grasping to the prosthesis itself, we can restore intuitive and dexterous control,” George said. “Simple tasks become simple again.”
“For this study, the machine controller did not adapt in real-time since the experiments were relatively short in duration,” Trout said in an email to The Debrief. “We didn’t want the machine algorithm to be changing while the user was still learning to control the hand. One can imagine how difficult it could be to learn to use a new hand that is constantly changing its behavior.”
Looking ahead, the Utah NeuroRobotics Lab’s goal is to integrate these intelligent sensors with implanted neural interfaces so users can control their prosthesis directly with their thoughts while regaining a sense of touch.
“In the future, the algorithm could indeed adapt in real-time as the user uses the bionic hand more and in different ways,” Trout said. “This is one of the future directions we would like to explore at the Utah Neurorobotics Lab.”
“In theory, the AI model that is attempting to assist the user in real-time could continuously learn from data while the bionic hand is being used,” Trout added. “What we would hope to see is that the autonomous agent be able to provide better assistance to the user as it is presented with a greater variety of scenarios, and the user continues to learn alongside the AI to leverage its improved functionality.”
The study, “Shared human-machine control of an intelligent bionic hand improves grasping and decreases cognitive burden for transradial amputees,” was published on Dec. 9 in Nature Communications.
Chrissy Newton is a PR professional and the founder of VOCAB Communications. She currently appears on The Discovery Channel and Max and hosts the Rebelliously Curious podcast, which can be found on YouTube and on all audio podcast streaming platforms. Follow her on X: @ChrissyNewton, Instagram: @BeingChrissyNewton, and chrissynewton.com. To contact Chrissy with a story, please email chrissy @ thedebrief.org.