Breath Powers Pediatric Prosthetic Hand

Virtual Prototyping Produces Intuitive Device for Children

One simple breath and the prosthetic hand closed. Another breath opened the fingers, the movement happening in a heartbeat. “It feels like magic,” a child with an upper limb difference in Bengaluru, India, told the University of Oxford team that made the prototype.

Called Airbender, the wearable upper limb prosthetic is designed to be affordable, comfortable, and intuitive for young users. A unique turbine and gearbox optimize torque and revolutions per minute (RPM).

“Body-powered prosthetics only considers the musculoskeletal system,” observed Jeroen Bergmann, an associate professor of engineering sciences who leads the project at the University of Oxford. “I came up with the idea for a breath-powered system after patients and clinicians expressed a need for new devices that are easy to use, especially for children.”

The prosthetic limb with the breath-powered hand attached. Attached to the hand is the Tesla turbine and breathing tube.

The Airbender wearable upper limb prosthetic. (Image credit: University of Oxford)

Currently, the most prevalent prosthetic hands rely on technology that hasn’t changed much in more than 200 years. These cable and harness systems require muscular strength to operate.

To develop a breath-powered system, Bergmann recruited postdoctoral biomedical engineering researcher Dr. Vikranth Nagaraja for the project in February 2019. A year later, postdoctoral engineer Dr. Jhonatan Da Ponte Lopes joined them.

To produce an operational device, Bergmann also partnered with the nonprofit organizations Mobility India and St. John’s Medical College Hospital in Bengaluru, U.K.-based charity LimbBo Foundation, and 3D LifePrints. The team received invaluable manufacturing support from the departmental workshops, particularly 3D-printing support from Dr. Peter Walters.

“The clinical and patient communities helped us tremendously,” Bergmann said. “Now we are optimizing the hand functionality and manufacturability and will be exploring new funding opportunities to move our work into the clinical testing phase.”

Shrugging Off the Past

Worldwide, 40 million people have limb differences, an inclusive term for body parts missing congenitally or lost (e.g., from trauma). Upper limb differences can range from missing part of a finger to both arms. Nagaraja noted that according to estimates from published literature, only about 3% to 5% of patients with limb differences receive prosthetic care.

“There’s evidence showing that children who lose limbs and are fitted with prosthetics have a higher level of development than those who don’t get a prosthetic at all,” Da Ponte Lopes pointed out.

“There’s evidence showing that children who lose limbs and are fitted with prosthetics have a higher level of development than those who don’t get a prosthetic at all.”

In 1818, a German dentist named Peter Baliff introduced the first upper limb system that enabled the wearer to flex the fingers on a prosthetic hand through trunk and shoulder movement. World Wars I and II prompted advancements such as sockets and tool attachments.

By the late 1940s, a unique cable mechanism began replacing the straps. Originally patented by Irish inventor Ernest Monnington Bowden in 1898 for bicycles, the cable consists of a thin wire enclosed in a larger outer sheath. Shoulder shrugging pulls and releases a gripper attached to the Bowden cable, opening and closing the prosthetic hand or hook in proportion to that tension.

“But these systems have been reported to be very inefficient,” Nagaraja said. “The wearer ends up exerting almost twice the amount of force to get a good grip. It’s tiring.”

Additionally, children with a congenital upper limb deficiency have been found to have lower strength in both their deficient and intact arms compared to nondisabled children. These children often lack adequate muscular strength to operate these devices comfortably.

Hot and humid conditions make it even more challenging to use a harness-based system. Children must also be refitted two or three times a year to accommodate for their constant growth, which insurance companies often will not cover. The prosthetic alone can cost $3,600 to $12,000 or more, making it unaffordable to low-resource communities that often rely on out-of-pocket expenditures for health.

While weighing alternatives, Bergmann reasoned that humans intuitively use breathing to coordinate and control functions such as speech, singing, and even playing woodwind and brass musical instruments. He sought to harness respiration to both power and guide assistive technology.

Building a Prototype

Bergmann’s first breath-powered device helped him formulate key questions, but admittedly didn’t resemble a real prosthesis. The cable and pulley system on a frame with four 3D-printed fingers weighed half a kilogram (slightly over 1 pound). The fingers took nearly a minute to close fully.

Extracting power from air entering a device calls for a turbine, but the high humidity and food and other debris in our breath can easily damage a turbine’s blades. Bergmann realized that wouldn’t be a problem with a Tesla turbine.

In 1913, famed inventor Nikola Tesla patented a bladeless turbine that uses disks mounted on a shaft. Fluid goes between the slightly spaced-out disks, making the rotor spin. However, when Tesla scaled up the turbine, high speeds damaged the disk material at the time, lowering the overall mechanical efficiency.

Prosthetic components, including the hand with a wrist adapter to attach to the arm prosthetic, the Tesla turbine, and the rubber tube from the turbine to the input nozzle/mouthpiece.

Components of the breath-powered prosthetic. (Image credit: University of Oxford)

Tesla’s invention couldn’t replace the piston combustion engine, but his concept suited the Airbender project. “It is bidirectional in nature so you can use two tubes for different movements instead of two turbines or for introducing a complex mechanism to reverse opening and closing,” Da Ponte Lopes explained. He added that the components’ simple geometries enabled them to be 3D-printed, keeping costs low.

He found extensive scientific literature on the airflow’s behavior between Tesla turbine disks, including the torque generated. One paper in the series presented a set of ordinary differential equations. Da Ponte Lopes reproduced those results with MATLAB®.

Modeling and Simulating the Virtual Prototype

Virtual prototyping and modeling with MATLAB® and Simulink® were essential to iterating through potential solutions for the prosthesis turbine. Bergmann’s team set up a virtual prototype using a system of ODEs (ordinary differential equations) that describe the fluid behavior inside the rotor.

“The libraries are one of the main advantages of MATLAB. For differential equations, you don’t have to spend time wondering whether the code is right,” said Da Ponte Lopes. “Another is visualization. I could make changes to the code very easily.”

“MATLAB and Simulink have well-developed functions and packages, which allow for rapid, high-quality design and testing of models.”

Bergmann agreed. “MATLAB and Simulink have well-developed functions and packages, which allow for rapid, high-quality design and testing of models,” he said.

What’s more, the pandemic forced the engineers to work remotely from various places. For a long time, they plugged away at virtual prototyping and computational fluid dynamics. Finally, with restrictions easing, Da Ponte Lopes returned to campus to build a testbench for characterizing and optimizing the Tesla turbine system.

“That’s not something you can buy, even if you have a million pounds to spend,” he said. “We needed to start from scratch.” Controlling the airflow amount and RPM in the testbench was crucial. Although the team had a specially designed mass flow controller that provided exact air amounts, RPM still posed a challenge. 

Tesla turbine system schematics. First, the turbine with the upper lid, rotor, and casing. Next, the rotor showing the bearings and discs. Finally, the casing containing the rotor, plenum chamber, fluid inlets/outlets, and tube inlet.

Design of the Tesla turbine system: a) Top-level components b) Rotor c) Fluid regions. (Image credit: University of Oxford)

A brushless DC motor could act like a brake or generator, depending on how the turbine functioned, similar to the brakes in Formula 1 cars. But Da Ponte Lopes wanted to ensure this expensive equipment would work before purchasing it, so he modeled the motor and control setup using Simulink first.

Optimizing the Prototype

The Airbender team wanted to ensure that users, especially children, could easily operate their prosthetic devices through variations in their natural breathing, rather than through bigger and harder breaths. To achieve this, they connected the Tesla turbine to a gearbox that converts airflow from each breath into a low speed with high torque. They opted for a gearbox with a worm gear that can change directions while staying locked in place through friction. This gearbox actuates the fingers.

“We had to be efficient and optimal in our designs to extract the most power from a child’s breathing.”

Demonstration of the breath-powered prosthetic. (Video credit: University of Oxford)

“We had to be efficient and optimal in our designs to extract the most power from a child’s breathing,” Nagaraja remembered. “Any friction or losses would not have allowed us to achieve the needed functionality. That was weighing on our minds.”

With the testbench ready in July 2021, the engineers found that individual components worked separately, but not as an integrated whole. Fingers jammed, the gearbox wouldn’t spin, and other issues stemming from misalignments cropped up.

“Finally, after many iterations and fine-tuning, we had a breakthrough. At one point, I blew directly into the motor, and the fingers closed very quickly,” Nagaraja said. “We had a couple of children try this, and they were able to control the device the same way.” The international peer-reviewed journal Prosthesis published an overview of the Airbender system on July 29, 2022.

Lightening the Load

Every year in India, there are more than 40,000 new upper extremity amputations, mainly due to electrical burns, industrial accidents, or trauma, according to the nonprofit, Mobility India. The options for amputees in the country are usually ineffective or too expensive.

A child with an upper limb difference who doesn’t receive prosthetic care has asymmetrical loading on the trunk, which can affect gait, cause overuse injuries, and lead to scoliosis. Thwarted developmental growth hampers the child’s quality of life, and the ensuing problems ripple throughout the larger community. 

“Our hope is to create solutions for those who previously did not have any available. Apart from providing function, these devices empower children to interact with the external world and realize their full potential.”

Since Mobility India began in 1994, the organization has worked at the grassroots level to provide services for people with disabilities and other vulnerable groups. “Upper extremity assistive products remain limited locally to one type that only comes in two sizes,” said Ritu Ghosh, academics director and principal of Mobility India’s Rehabilitation, Research & Training Centre in Bengaluru.

“When we discussed the concept of the Airbender project, we felt this could be another design choice offered in the country,” she said. “The simplicity of the technology makes it appropriate not only for urban populations but for rural populations as well.”

Now the terminal device looks much more like a prosthetic hand. It fits a small wrist adaptor on a socket that clinicians produce for patients, forming a naturalistic interface. The engineering team anticipates that it could be manufactured at a cost comparable to traditional body-powered devices.

In 2022, the project team collaborated with Mobility India to recruit 15 children and teens with upper limb differences in Bengaluru for a clinical usability study of the latest Airbender prototype.

“It’s a collaborative and user-centered approach where we involve patients, clinical teams, and parents,” Nagaraja said. “We received favorable feedback. Some of the children were adventurous and playful with the device. Several teenagers seemed shy about breathing into the device in public settings, but said they liked the technology.” 

Afterward, the team published another peer-reviewed journal article that details the user testing. They plan to carry out further product development prior to entering clinical trials. Based on feedback, they intend to make improvements to the breath-powered device’s weight, sound, and appearance before bringing it to market.

Bar chart showing the median response of satisfaction for color, appearance, usefulness, reliability, comfort, and overall. Shape and noise received neither satisfied nor dissatisfied. Weight received a median response of unsatisfied.

Clinical usability study results. (Image credit: University of Oxford)

“Our hope is to create solutions for those who previously did not have any available,” Bergmann said. “Apart from providing function, these devices empower children to interact with the external world and realize their full potential.”

Read Other Stories

Panel Navigation


Machine Learning Aids Neurosurgeon During Brain Surgery

Improving Parkinson’s Surgery with AI and Signal Processing

Panel Navigation


Autonomous Pediatric Exoskeleton Takes Its First Steps

Robotic Medical Device Advances Physical Therapy for Children

Panel Navigation


Combining Artificial Intelligence and Footwear Improves Diabetes Treatment

Smart Shoe for Diabetic Neuropathy Gets a Tech Upgrade