Controlling drones with the wave of a hand

While a straightforward concept, the development process was far from simple. Headed by Frank Clemens from Empa’s Laboratory for High-Performance Ceramics, the sensor is made of piezo-resistive fibers incorporated into a wristband that registers the hand’s movements.

The electroconductive piezo-resistive fiber recognises a change in shape and converts this motion into an electrical signal, which can then be read by a terminal device and interpreted accordingly. In other words, drones and other robotic devices can be moved with basic motions such as the point of a finger.

While motion sensors have been around for many years, the Empa version takes this classic technology in a fundamentally different direction. Up until now, movements were typically recorded through visual sensors, such as cameras, as well as accelerometers and gyroscopes, in the case of rotational movements.

The trick is that this way of recording movement generally requires big, clear motions within a specific speed range – motions that are by and large unnatural for humans. The new Empa sensor, however, responds to even the smallest natural movement, even when made “off the cuff.”

Yet as Clemens notes, the Empa device couldn’t have existed without its forerunners. "It takes a combination of different sensors to develop new concepts,” he says. “Only then can we spot and use movements that weren't detectable with previous technologies."

In this case, it’s the combination of acceleration, rotation, and orientation sensors with the specialised fiber sensor that enables new commands to control electronic devices. Drones are an obvious application, but the Empa sensor could even be applied to open a garage door.

While in the prototyping stages, Clemens and his colleague Mark Melnykowycz experimented with attaching the piezo-resistive fibers to a piece of fabric. Although successful, the fabric was insufficient for using the sensor on the desired scale.

That’s where 3D printing came in, Clemens notes. “With the aid of additive manufacturing, we managed to integrate the sensor structure in non-textile materials," he says. The sensor could now be used in something as structural, yet basic, as a wristwatch.

In its current form, the Empa sensor is integrated into a conventional wristwatch strap to prevent restriction and obtrusiveness for the user. A regular bracelet is also conceivable, as it would achieve similar effects.

With the help of Swiss companies STBL Medical Research AG and Idezo, Clemens and his team then programmed the sensor to control a drone with specific hand movements. In addition to individual movements, the Empa sensor can also respond to movement sequences. Clenching your first twice quickly, for example, would be a different command than clenching it once slowly and once quickly.

Currently, the algorithm is being optimised at Bern University of Applied Sciences, with the goal of responding to even simpler gestures. Further research at ETH Zurich indicates that even the wristband may be a thing of the past too: the piezo-resistive sensor could also feasibly be contained in a plaster.

Although the project is still in its infancy, Empa researchers have wasted no time in sussing out potential applications and buyers. "Together with our industry partner STBL Medical Research AG, we are currently discussing a potential industrial implementation with partners from various sectors," Clemens says.