Knee-d a Lift?

Physically demanding work, of the sort often found in construction and manufacturing, takes a toll on workers. Repetitive lifting and carrying tasks, in particular, make these individuals susceptible to excessive fatigue. And when workers are heavily fatigued, they tend to stray away from using proper lifting techniques. This is especially true when they have to maintain a fast pace, like when keeping up with a conveyor belt. Ultimately, this fatigue results in higher incidences of injuries and reduced levels of productivity.

An innovative way to deal with this problem involves the use of back exoskeletons. These assistive devices utilize springs or motors to provide their wearers with a boost while lifting. However, they also require their users to lift with their back, which is otherwise an unsafe lifting technique. They can also be very cumbersome, and must be adjusted or removed before the wearer can perform any tasks aside from those that the exoskeleton was explicitly designed for.

Engineers at the University of Michigan have designed a new type of exoskeleton that is much more versatile and natural to use than existing back exoskeletons. Their device attaches at the knee, and it provides a boost to the quadriceps muscles. This allows users to lift with good form and move normally even when they are not performing a lifting task. Intelligence was also built into the system such that it can rapidly adapt to the amount and type of assistance that is needed.

The exoskeletons are built around off-the-shelf knee braces and electric motors scavenged from drones. The motors are geared in such a way that the wearer can walk with a natural gait, and assistance can be given by the motors as needed. Sensors measure factors like the angle of the knee joint, the orientations of the thigh and lower leg, and the force registered in the wearer’s shoe up to 150 times per second.

Data from the sensors is fed into a machine learning algorithm that predicts the best way to assist the wearer of the exoskeleton. This is in contrast to many existing systems, which can only assist in a limited set of tasks by following predefined programming. As a result, users of the new exoskeleton can do far more, and the system will rapidly adapt to whatever their needs may be at the moment.

The prototype device was tested in a small study of ten participants. Over the course of two days, they were instructed to use the exoskeleton while performing a variety of tasks both when fresh and when fatigued. The participants generally reported that they were very satisfied with the assistance received, and as the workload increased, so did their level of satisfaction. The study also showed that the device significantly improved the speed at which users can carry out their tasks.

At present, the exoskeletons cost about $4,000 to produce, but the team thinks that cost could be halved if they were produced at scale. In any case, that could prove to be a small price to pay for the reductions in injuries and increases in productivity the system can offer.