The new hybrid robotic hand blends soft and rigid parts with touch-sensitive technology, allowing for precise and flexible object handling. (Credit: Sriramana Sankar/Johns Hopkins University)
Advanced prosthetic grasps objects with control, detects textures
In a nutshell
- A new prosthetic hand combines a rigid internal skeleton with soft, flexible joints, mimicking the structure of a real human hand for both strength and delicate touch.
- A multilayered sensor system in the fingertips enables the prosthetic to distinguish textures and objects with 98.38% accuracy, significantly outperforming traditional prosthetics.
- While designed for sensory feedback, the system has not yet been tested on amputees, but researchers aim to refine the technology to restore a more natural sense of touch.
BALTIMORE — When someone loses a hand, today’s prosthetic options force painful compromises. Rigid prosthetics offer strength but can’t handle delicate objects gently. Soft robotic alternatives provide gentleness but lack gripping power. And neither option lets users actually feel what they’re touching.
A breakthrough from Johns Hopkins University researchers aims to finally solve these problems. In a newly published study in Science Advances, the research team has developed what they’re calling a “natural biomimetic prosthetic hand” that blends rigid and soft materials while adding touch-sensing abilities based on human skin.
“The goal from the beginning has been to create a prosthetic hand that we model based on the human hand’s physical and sensing capabilities—a more natural prosthetic that functions and feels like a lost limb,” says lead study author Sriramana Sankar, a Johns Hopkins biomedical engineer, in a statement. “We want to give people with upper-limb loss the ability to safely and freely interact with their environment, to feel and hold their loved ones without concern of hurting them.”
Best of Both Worlds
Instead of choosing between rigid or soft designs, the research team took inspiration from human anatomy. Our hands combine rigid bone structures with soft tissues and joints, so why not do the same with prosthetics?
The Johns Hopkins team built a hand with a hard 3D-printed internal skeleton surrounded by soft, independently controlled joints made of silicone. But their biggest innovation might be the touch-sensing system built into the fingertips.
The researchers embedded three different types of sensors within the prosthetic fingertips to mimic how human skin works. Our skin contains specialized cells called mechanoreceptors that detect different aspects of touch, from light pressure to vibrations to skin stretching. The artificial version includes layers of sensors that work together to create a rich picture of whatever the hand is touching. The system converts touch data into patterns similar to the electrical signals our nerves would normally send to our brains.
Putting It to the Test
In lab tests, the hybrid hand showed remarkable abilities. When asked to identify 26 different textured surfaces, from smooth plates to various ridged patterns, it achieved 98.38% accuracy, far outperforming both purely soft robotic fingers (82.31%) and rigid prosthetic fingers (83.02%) tested with the same surfaces.
Forearm muscle signals control the new hybrid robot hand, effortlessly gripping and adjusting to different objects. (Credit: Sriramana Sankar/Johns Hopkins University)
The hand was also tested with 15 everyday objects including stuffed toys, fruit, dishes, and water bottles. It correctly identified these items with 99.69% accuracy while handling them appropriately; gentle with delicate items, firm with heavier ones.
Perhaps most impressive was when the hand picked up a thin plastic cup filled with water using just three fingers without crushing or denting it, a task that would be nearly impossible for conventional prosthetics.
“We’re combining the strengths of both rigid and soft robotics to mimic the human hand,” says Sankar. “The human hand isn’t completely rigid or purely soft—it’s a hybrid system, with bones, soft joints, and tissue working together. That’s what we want our prosthetic hand to achieve. This is new territory for robotics and prosthetics.”
How It Works
The prosthetic uses electromyography (EMG), the same control method used in many modern prosthetic hands. EMG sensors detect electrical signals from remaining muscles in the user’s arm, allowing them to control the hand’s movements by intentionally flexing those muscles.
“This hybrid dexterity isn’t just essential for next-generation prostheses,” says study author Nitish Thakor, a Johns Hopkins biomedical engineering professor. “It’s what the robotic hands of the future need because they won’t just be handling large, heavy objects. They’ll need to work with delicate materials such as glass, fabric, or soft toys.”
A key benefit of the hybrid design is efficiency. It generates three times more gripping force than a purely soft robotic hand while needing only a quarter of the air pressure to operate. The hybrid hand produced 1.8 Newtons of force at just 7 psi (pounds per square inch), compared to 0.55 Newtons at 28 psi for a soft robotic hand.
Restoring the Sense of Touch
The bionic hand grips a ball. (Credit: Sriramana Sankar/Johns Hopkins University)
Beyond just improving grasping abilities, this technology might eventually restore the sensation of touch to prosthetic users. While the current study focused on demonstrating the hand’s physical capabilities, the researchers designed the system with sensory feedback in mind.
“If you’re holding a cup of coffee, how do you know you’re about to drop it? Your palm and fingertips send signals to your brain that the cup is slipping,” says Thakor. “Our system is neurally inspired—it models the hand’s touch receptors to produce nerve-like messages so the prosthetics’ ‘brain,’ or its computer, understands if something is hot or cold, soft or hard, or slipping from the grip.”
This technology builds on the lab’s previous work, which included creating the world’s first electronic “skin” with human-like pain sensing in 2018. While the system is designed to provide sensory feedback, it has not yet been tested on amputees to determine how effectively users perceive and respond to the touch signals.
Looking Forward
The current prototype’s 1.8 Newtons of gripping force, while an improvement over soft robotic hands, falls well short of a human finger’s capability (around 32 Newtons) or traditional rigid prosthetics (about 34 Newtons). It also relies on an air compressor to function, which would be impractical for everyday portable use.
For people who’ve lost hands, this research offers a glimpse of prosthetics that might one day feel like a genuine replacement that is able to handle both fragile and heavy objects, sense textures and shapes, and respond naturally to the user’s intentions. Prosthetic technology has long focused on looks and basic functionality, but this approach aims to restore what matters most: the hand’s remarkable ability to both act and feel.
Paper Summary
Methodology
Researchers combined soft silicone (Dragon Skin 10) with rigid 3D-printed polylactic acid structures. Each finger contains three pneumatic joints working between rigid skeletal components, mimicking human anatomy. The fingertips house three sensing layers: outer and middle piezoresistive fabric sensors and an inner piezoelectric sensor. The system processes touch information using neuromorphic encoding to mimic human nerve signals. Testing involved mounting the hand on a robotic arm and comparing its performance with both soft robotic and rigid prosthetic fingers.
Results
The hybrid design achieved 98.38% accuracy in distinguishing between 26 textured surfaces versus 82.31% for soft fingers and 83.02% for rigid fingers. When grasping and identifying 15 everyday objects, the hand reached 99.69% classification accuracy. It generated three times more force (1.8 Newtons) than soft robotic hands while using one-quarter of the pneumatic pressure (7 psi versus 28 psi). The multilayered sensing approach showed that each layer contributed unique information, with combined layers significantly improving overall performance.
Limitations
The current prototype generates only 1.8 Newtons of force, far below a human finger’s 32 Newtons. It requires external pneumatic actuation (air compressor and valves), making it impractical as a portable prosthetic. The sensory feedback capability, while designed into the system, wasn’t directly tested with amputees. The hand has fewer degrees of freedom than human hands and lacks long-term durability testing under real-world conditions.
Discussion and Takeaways
The study demonstrates that combining soft and rigid components yields better performance than either approach alone. The multilayered sensing system significantly improves texture and object identification—crucial for users who can’t rely solely on vision. The neuromorphic approach suggests potential for providing natural sensory feedback through nerve stimulation in future iterations. This hybrid design philosophy could influence both prosthetics and general-purpose robotics, especially for applications requiring both strength and delicate touch.
Funding and Disclosures
This research was funded by the Department of Defense through the Orthotics and Prosthetics Outcomes Research Program (grant W81XWH2010842) and the National Science Foundation. The researchers declared no competing interests that would influence their findings. All experimental data has been made publicly available through a permanent Zenodo repository.
Publication Information
The paper, “A natural biomimetic prosthetic hand with neuromorphic tactile sensing for precise and compliant grasping,” was authored by Sriramana Sankar, Wen-Yu Cheng, Jinghua Zhang, Ariel Slepyan, Mark M. Iskarous, Rebecca J. Greene, Rene DeBrabander, Junjun Chen, Arnav Gupta, and Nitish V. Thakor. It was published in Science Advances (Volume 11, eadr9300) on March 5, 2025.
