The Swiss Federal Institute of Technology Zurich held its second “Cybathlon” last year – a tournament showcasing life-changing technologies for people with disabilities.
The Cybathlon is an Olympic-style international event for bionic athletes. Think of the Paralympics complete with cyborgs, AI and robotics.
Last year, 51 teams worldwide competed against each other in the second ever event, organized by the Swiss Federal Institute of Technology in Zurich (ETHZ).
The aim of the competition is to stimulate the development of technologies to improve the daily lives of people with disabilities. Teams can participate in six disciplines:
powered leg prostheses, powered arm prostheses, powered exoskeletons, electric wheelchairs, paraplegic bicycles that use their paraplegic pilot’s own electrically stimulated muscles, and a race between the brain and the computer where competitors must control avatars in a software game by brain power alone.
Competitive teams range from large corporations to university labs and one-man bands, handling cutting-edge prototypes and commercially available products. The winners will receive two medals:
one for the technologists, the other for the human pilots who have been an integral part of the development of the devices and trained long and difficult to learn new skills.
The event was originally scheduled to take place in Zurich in May 2020, but was postponed to November 2020 with teams competing virtually on identical courses around the world due to Covid-19 travel restrictions.
Aldo Faisal, Professor of Artificial Intelligence and Neuroscience at the Brain and Behavior Lab, Imperial College London, has been the captain of the Imperial team since 2014.
user at the heart of their designs. If they don’t, they will fail. This year, our teams are all embracing AI to enable or restore people’s interaction with their environment, and to change those lives for the better.
It is a real opportunity for them to learn that there is more to AI than analyzing and processing data flows to meet the needs of large companies. ”
In 2020 Imperial set up teams in three disciplines: bicycles for paralyzed cyclists, the robotic arm and the electric wheelchair.
In any case, the teams must complete a series of mundane tasks that are difficult for people with different disabilities, and then challenge what is possible with today’s technology.
London-based Canadian Conrad Bona was born without a left hand in the 1960s. “I’ve had a prosthesis since I can remember; I had my first mechanical hand at the age of 16. I am now 51 years old and the pilot of the prosthetic hand challenge in the Cybathlon 2020 for Team Imperial.
” Bona loves to be part of the Imperial College development team led by Irene Mendez Guerra, a PhD student and bioengineer who works in the neuromechanics and rehabilitation technology laboratory.
“Our prosthesis is unique,” she explains. “It is designed to tackle everyday but difficult real-world tasks, such as using scissors, and to significantly improve the quality of life for our users.
” Missing a limb, especially an arm, can have a devastating effect on the quality of a person’s life. Twenty thousand different people of the upper limbs are under treatment and rehabilitation in the NHS, but many of the prostheses currently on the market are non-intuitive, rudimentary, and require lengthy training to be used effectively.
From the start of the project, the Imperial team wanted to design a robotic prosthesis that the wearer could quickly and easily learn to use.
The functioning of the team’s hand and wrist prosthesis is controlled by muscle activity in the remaining muscles in Bona’s upper arm.
When muscles contract, they generate a difference in tension that can be detected by sensors on the skin’s surface. Irene’s team uses signal processing and machine learning to decode the muscle signals generated by Bona’s brain when he thinks about moving his missing hand.
These decoded signals in turn generate corresponding commands for the motors in the prosthesis. Conversely, signals return from the robotic hand to Bona’s brain from sensors in the fingertips, giving him haptic feedback or a coarse sense of touch, allowing him to distinguish between objects of different hardness and shape.
The haptic feedback mechanism also helps to improve the quality of the bionic handle. The prosthesis is built from the ground up at Imperial – all robotics, sensors and electronics.
“The tight integration of each element makes it compact and stable, and Conrad was there at every iteration to provide real user feedback,” says Guerra. Without fine control and without intuitive interfaces, the operation of many prostheses is a frustrating exercise for the user.
As many as 40 percent of people with disabilities who have received prostheses leave them within six months. According to Professor Faisal, this high adoption failure rate could be the result of a design process that builds shiny robotics in a lab that is separate from the real needs of its users.
“The Cybathlon competition tries to rectify this by involving disabled pilots at all stages of development,” he added. Bona confirms, “It’s all about the user interface: if it’s not intuitive, people won’t use it.
Since the competition hand is a one-off prototype and not commercially available, it’s significantly better than what I use every day. functioning robotic hand. With this the level of control is quite phenomenal.
Picking up fine objects like a key or a credit card and putting them in a lock or a slot, I would always do that with my natural hand, but with this I can use the prosthesis.
hard to leave when I leave the lab! ”At the Cybathlon event, Bona had to face everyday challenges with his robotic hand, such as hanging laundry on a line with clothespins, cutting a baguette, and picking up and screwing on a light bulb.
Fifteen minutes before the start of their match, the Imperial ARM thumb’s articulation broke and the team had to compete with an earlier prototype, which did not perform well enough to secure a medal.
In the electric wheelchair race, Mahendran Subramanian, based in Imperial’s Brain and Behavior lab, the biomedical engineer behind an extraordinary wheelchair, is almost as equipped with sensors as a top-of-the-line Tesla.
A whole new drive-by-eye system allows severely disabled people to navigate their surroundings using a self-driving electrically powered wheelchair, with just one look.
It’s as simple as staring at the intended destination and confirming it with a wink. It only takes a few minutes for first-time users to learn how to operate the chair, providing safe independent mobility for people so severely disabled that they cannot use conventional joystick or other navigation interfaces such as tablets and touchscreens.
“We’ve used AI and machine learning to decipher the difference between a focused look and a look. You wouldn’t want to end up with a wall you looked at instead of the kettle you wanted to go to make a cup of tea, ‘says Subramanian.
The pilot of the extraordinary wheelchair is Paul Moore, a business owner with a spinal cord injury, whose Activelinx company specializes in assistive technology for the disabled.
An infrared gaze tracker tracks Moore’s eye movements to determine where he wants to go in the physical world. A simple RGB camera with built-in depth sensor calculates the distance to its desired destination.
As soon as he confirms his intention with a wink of three seconds, the software combines the destination “kettle in the kitchen” with commands sent to the navigation software.
A cup of tea is coming. Borrowing technology from self-driving vehicles, a two-dimensional lidar sensor delivers environmental data to a Simultaneous Localization and Mapping (SLAM) algorithm on a laptop mounted on the back of the wheelchair.
The SLAM creates a map and locates the wheelchair in it, in real time, allowing the wheelchair to detect obstacles or objects moving in the vicinity and plan a route to the kettle to avoid them safely.
“We deliberately used inexpensive ready-to-use sensors to keep the costs of our conversion kit low. That makes it more accessible to the people who really need it, ”says Subramian.
“Ultimately, we plan to market it as a kit to modify existing power wheelchairs for about the same price as it costs to add a joystick.” “This drive-by-eye technology will be revolutionary for people with severe neurological disease,” explains Moore.
“The Cybathlon is more than a competition; it is a phenomenal showcase for future technologies. We are all very strongly convinced that this should be within the reach of every disabled person.
There’s no question that the drive-by-eye concept is amazing and could be transformative for some people in the future, but it still has a way to go before we see it in the market.
“Subramian’s seat design has no extra rails – so like any regular chair, it can’t climb stairs. This meant it couldn’t have won the Cybathlon race because the track included a flight of stairs.
” But our prototype is more practical than any other vehicle, “Moore notes.” It fits in a car or traveling on the subway. Most wheelchair users are not looking for the equivalent of a four-wheel drive or an SUV. We just want to be able to travel independently.
Net like anyone else. ”Due to the travel restrictions of Covid-19, Subramian and Moore had to compete virtually at the Allia Business Center in Peterborough, instead of Zurich, but unfortunately the limited start-up time meant that the team had to withdraw at the last minute.
The third event in which Imperial participated is the Functional Electrical Stimulation (FES) cycling race. FES is a technique that has been around for more than two decades. It allows riders with a spinal cord injury to use their paralyzed legs for cycling by electrically stimulating the muscles to contract.
The team uses a modified BerkelBike recumbent bike as a platform, invented by the Dutch chemist and exercise scientist Rik Berkelmans, who explains:
“The combination of FES and cycling provides enormous benefits in terms of improvements in life expectancy, reversal of muscle atrophication and can help of the other health problems associated with debilitating spinal injuries, such as obesity and cardiovascular problems.
Imperial’s driver for this race is Johnny Beer, whose life changed at the age of 16 when he was paralyzed with a complete spinal cord injury in a trampolining accident. says the passionate determined and competitive cyclist who trains several times a week.
Bear has been riding a BerkelBike since 2012. The event would normally be held on a bike path, but Covid-19 forced the FES racing teams to put their bikes on Wahoo V4 standardized training roles.
used by cyclists around the world to train during bad times.Weather. Each roll was checked and calibrated at the Cybathlon headquarters in Zurich before it was shipped.
The move inwards meant any aerodynamic improvements to the bikes negated by the lack of wind resistance Self-adhesive electrodes stimulate Bee’s muscles r and a computer monitors the pedal position so that the correct muscles can be stimulated during the 360 degree pedaling cycle.
Thai national Nat Wannawas is Imperial’s robotic engineer behind the FES entry, and his job was to fine-tune that process using AI and machine learning to deliver an optimal level of stimulation given Beer’s speed and fatigue levels.
The combination of training and tuning the software resulted in Imperial’s only medal at the 2020 Cybathlon. Beer and Wannawas won the silver medal and lost it to the Dutchman Sander Koomen.
The next Cybathlon will be held in 2024, once again expanding the capabilities of prosthetics to the next level and showcasing life-changing technologies for people with disabilities.