According to the World Health Organization (WHO), an estimated 424,000 people across the world die each year from falls. A total of 37.3 million falls are severe enough to require medical attention. Of course, adults older than 65 suffer the greatest number of fatal falls.
Researchers at Scuola Sant’Anna in Italy and EPFL (Ecole polytechnique federale de Lausanne) in Switzerland are looking to prevent elderly people from falling with a new smart exoskeleton.
The Active Pelvis Orthosis (APO), an exoskeleton worn from the waist down, recognizes in just 350 milliseconds that a person is about to fall and applies force to help the legs counteract the slippage.
The researchers claim in Nature Health Sciences the APO is the first exoskeleton to prevent people from falling. Exoskeletons typically assist or enhance regular movement.
The exoskeleton is made of carbon fiber, so it’s lightweight at just 11 lbs, and it has motors at the hips. It uses an algorithm to learn the person’s gait in three minutes. To set up the APO, a user straps it on, programs their weight, and takes three steps.
The algorithm detects diversions from that gait, which indicates a loss of balance, and applies force to the hips to keep the person upright.
The system was tested on eight people, who had an average age of 68.9, walking on a custom treadmill. To simulate a foot slipping, the right or left half of the treadmill would jolt outward.
One of the test patients was 69-year-old Fulvio Bertelli. He says the APO made him “feel more confident.” Of course, the next step is to test the APO in more situations where falls are more realistic and not simulated on a treadmill.
“This work paves the way for imagining a completely new generation of exoskeletons that will actually be effective outside of research laboratories thanks to their ability to augment users’ movement and make their mobility more stable and safe,” says Scuola Sant’Anna professor Nicola Vitiello. “To reach these goals, exoskeletons must be endowed with features, like the one proved in this study, that really take into account what users can experience in real-life unstructured environments.”