A group of knee exoskeletons, created using commercially available knee braces and drone motors at the University of Michigan, has been discovered to effectively reduce fatigue during lifting and carrying activities. Researchers claim that these devices assist users in maintaining a proper lifting posture even when they are tired, which is crucial in preventing workplace injuries.
A group of knee exoskeletons, developed with commercially available knee braces and drone motors at the University of Michigan, has been found to effectively reduce fatigue while performing lifting and carrying tasks. Research indicates that these exoskeletons help users maintain a proper lifting posture, even when they are fatigued, which is essential for preventing injuries on the job.
“Instead of simply supporting the back and compromising proper lifting techniques, we focus on strengthening the legs to sustain the correct form,” stated Robert Gregg, a professor of robotics at U-M and the lead author of the study published in Science Robotics. “This approach is different from the existing practices in various industries.”
Workers in jobs that require frequent lifting, such as construction and manufacturing, often utilize back braces. Back exoskeletons, which employ springs or motors to assist with lifting, represent an emerging technology. However, these devices are designed with the assumption that unsafe lifting or bending is occurring, and they can be cumbersome, needing to be turned off for movements outside of lifting tasks, according to Gregg.
The research team from Michigan claims that their knee exoskeletons are the first to enhance the quadriceps muscles, which are crucial for effective squat lifting, providing a less obtrusive method to safeguard workers from back injuries. Participants in the study tested these exoskeletons while lifting and carrying a 20 lb kettlebell.
The activities involved included lifting the kettlebell from the ground and placing it back down, as well as lifting and carrying it across flat surfaces, inclines, and stairs. The study revealed that fatigued participants were able to maintain better posture with the exoskeleton’s help, and they also lifted faster, showing only a 1% reduction in speed compared to their pre-fatigue paces, whereas without the exoskeleton, they slowed down by 44%.
“This is particularly significant in scenarios where workers must keep pace with a conveyor belt. Typically, when a worker grows fatigued, they might maintain the same speed but at the cost of their posture, leading to increased risk of injury,” explained Nikhil Divekar, a postdoctoral research fellow in robotics at U-M and the first author of the study.
Participants reported a positive experience, as most expressed they were quite or very satisfied with the exoskeleton; their only exception was while walking on flat ground, where their satisfaction was moderate. This aligns with the minimal assistance required for quadriceps during such an easier task, with Gregg noting it provided just sufficient support to counterbalance the weight of the exoskeleton.
A key factor in the exoskeleton’s comfort is the motors and their gearing, which allows users to move their knees naturally. Additionally, the software is designed to predict the support needed by assessing the knee joint angle, the alignment of the thigh and lower leg, and the force detected by a sensor located in the user’s shoe.
By capturing these three metrics from both legs, the device can determine the user’s motion and the appropriate level of assistance required. These readings are taken 150 times every second, allowing the exoskeletons to adapt seamlessly to various tasks.
This method deviates from many existing exoskeleton controllers that follow predefined patterns limited to a specific range of activities. Such controllers can face challenges when tasks shift, sometimes taking a full second to ascertain what the user intends to do, Gregg noted.
“Imagine your exo is programmed for going upstairs, but you want to go downstairs; that could create an issue,” he remarked.
The innovative controller combines a physics model with machine learning, which helps prevent unexpected movements from the exoskeleton if the user behaves outside the trained activity range of the controller.
The prototypes developed in the lab are priced around $4,000 per pair. However, Gregg anticipates that if produced on a larger scale, the cost could drop to approximately $2,000 per pair.
The study involved ten participants, comprising five men and five women, who completed all activities on two separate days: one day while rested and another while fatigued. To induce fatigue, each participant performed a series of squat lifts with the kettlebell until they could no longer continue without significant breaks. All participants were familiar with proper squat lifting techniques.
The research was financed by the National Institutes of Health.
The team has sought patent protection with the help of U-M Innovation Partnerships and is looking for partners to commercialize the technology.
