These tiny drones help students learn about algorithms

Interacting with drones teaches students the importance of control algorithms, so they can go on to make aerial vehicles that can carry out complex tasks.

When Crazyflie drones go crazy, bachelor students learn how important proper programming is. Because otherwise those quadcopters, the size of a palm of a hand, crash. For real, not only in theory. “The interaction with an actual drone is very beneficial when learning about control theory,” says Paul Beuchat.

In five consecutive semesters the doctoral student at the Automatic Control Lab delivered the hands-on quadcopter control course to bachelor students at the ETH Department of Information Technology and Electrical Engineering. “Considering the control algorithm is the focus of the course, students develop a much stronger connection to the design of their algorithm when they have the opportunity to interact with its embodiment in the form of a Crazyflie”.

Students are being taught much more than just programming using components of the Crazyflie, a 32 gram, lightweight drone. Raffaello D'Andrea, Professor at the Institute for Dynamic Systems and Control at ETH Zurich, conjured up an entire light show combining custom algorithms, software, and hardware with some Crazyflie components, and surprised the visitors at a TED event.

“Due to their small load capacity and flight time, the application of Crazyflies is mostly restricted to being a swarm of lights or a swarm of sensors”, Beuchat admits. But he thinks they could be more widely used.

Quadcopters in general are very versatile, since they are mechanically relatively simple and at the same time very reliable. Their advantage: even if one of their rotors fails, they can still land safely. D'Andrea therefore uses various micro quadcopters for the intricate shows he designs with the ETH spin-off Verity Studios, including the one for Circus Knie and singer Celine Dion.

With their practical experience, the bachelor students hold the potential for being able to develop and programme drones for more complex missions in the future. For example, the omnidirectional service drone Voliro emerged from a student project and has been designed to complete maintenance tasks at height.