Scientists built an AI co-pilot for prosthetic bionic hands

Building an AI Co-Pilot for Prosthetic Bionic Hands

Modern bionic hand prostheses have made significant strides in terms of dexterity, degrees of freedom, and capability. However, despite these advancements, many amputees who try these advanced prostheses ultimately abandon them, never to use them again. According to Jake George, an electrical and computer engineer at the University of Utah, up to 50 percent of people with upper limb amputation give up on these prostheses.

The primary issue driving users away from bionic hands is their difficulty in controlling them. George explains that the main problem lies in the lack of autonomy in these prostheses. Unlike natural movements, which rely on an elaborate system of reflexes and feedback loops, bionic hands require users to think about every detail, from grasping a paper cup to catching a ball mid-flight.

The Micro-Management Issue

Bionic hands' control problems stem from their inability to mimic the natural reflexes and feedback loops that occur in the human body. When an object begins to slip, tiny mechanoreceptors in the fingertips send signals to the nervous system, which makes the hand tighten its grip. This reflex is just one of many ways the brain automatically assists in dexterity-based tasks.

Most commercially available bionic hands do not have this built-in autonomic reflex. Instead, everything must be controlled by the user, which makes them extremely involved to use. George likens this to trying to think about precisely adjusting the position of 27 major joints and choosing the appropriate force to apply with each of the 20 muscles present in a natural hand.

The Current State of Bionic Hand Control

Currently, users control bionic hands via an app where they can choose predetermined grip types and adjust forces applied by various actuators. A slightly more natural alternative is electromyography, where electric signals from the remaining muscles are used to command the bionic hand. However, this too is far from perfect.

Marshall Trout, a University of Utah researcher and lead author of the study, explains that grasping an object requires reaching towards it, flexing the muscles, and then concentrating on holding the muscles in the exact same position to maintain the same grasp. This level of concentration is not only tedious but also exhausting.

Building an AI Bionic Hand Co-Pilot

To address the issue of bionic hand control, George and his team started by fitting a commercially available bionic hand with custom sensors. The researchers replaced the fingertips with silicone-wrapped pressure and proximity sensors, allowing the hand to detect when it was getting close to an object and precisely measure the force required to hold it without crushing it or letting it slip.

To process the data gathered by the sensors, the team built an AI controller that moved the joints and adjusted the force of the grip. By repeating back-and-forth movements countless times, the team collected enough training data to have the AI recognize various objects and switch between different grip types.

Shared Control: The Key to Success

The brand-new touch the team applied was deciding what was in charge of the system. Earlier research projects that investigated autonomous prostheses relied on the user switching the autonomy on and off. By contrast, George and Trout's approach focused on shared control.

"It's a subtle way the machine is helping," George says. "It's not a self-driving car that drives you on its own, and it's not like an assistant that pulls you back into the lane when you turn the steering wheel without an indicator turned on." Instead, the system quietly works behind the scenes without it feeling like it's fighting the user or taking over.

Testing the AI Bionic Hand

To test their AI-powered hand, the team asked intact and amputee participants to manipulate fragile objects, such as picking up a paper cup and drinking from it, or taking an egg from a plate and putting it down somewhere else. Without the AI, they could succeed roughly one or two times in 10 attempts. With the AI assistant turned on, their success rate jumped to 80 or 90 percent.

The Next Steps

Trout emphasizes that the next step is to take this system into the real world and have someone use it in their home setting. So far, the performance of the AI bionic hand has been assessed under controlled laboratory conditions, working with settings and objects the team specifically chose or designed.

George cautions that this hand is not as dexterous or easy to control as a natural, intact limb. However, he believes that every little increment that we make in prosthetics allows amputees to do more tasks in their daily life. To get to the Star Wars or Cyberpunk technology level where bionic prostheses are just as good or better than natural limbs, we'll need more than just incremental changes.

The Future of Bionic Prostheses

Trout notes that we're almost there as far as robotics go. These prostheses are really dexterous, with high degrees of freedom. However, there's no good way to control them. This in part comes down to the challenge of getting the information in and out of users themselves.

The team is currently working on neural interface technologies and looking for industry partners. George emphasizes that the goal is to combine all these approaches in one device. "We want to build an AI-powered robotic hand with a neural interface working with a company that would take it to the market in larger clinical trials."

The future of bionic prostheses is exciting, and the possibilities are endless. With advancements in AI, robotics, and neural interface technologies, we can expect to see significant improvements in the control and functionality of bionic hands. As researchers continue to push the boundaries of what is possible, we may soon see the development of prosthetic limbs that are not only indistinguishable from natural limbs but also surpass them in terms of dexterity and capability.

Source: https://arstechnica.com/ai/2025/12/scientists-built-an-ai-co-pilot-for-prosthetic-bionic-hands/