The Micro Air Vehicle Laboratory (MAVLab) of Delft University of Technology has created the currently smallest autonomous racing drone in the world. It measures 10 centimeters in diameter and weighs 72 grams. It uses only a single camera and very little onboard processing in order to autonomously fly through a racing track with a speed that rivals that of the fastest, bigger autonomous racing drones.
Drone racing by human pilots is becoming a major e-sport. In its wake, autonomous drone racing has become a major challenge for artificial intelligence and control. Over the years, the speed of autonomous race drones has been gradually improving, with some of the fastest ones in recent competitions now moving at 2 m/s. Most of them are equipped with high-performance processors, with multiple, high-quality cameras and sometimes even with laser scanners. This allows for state-of-the-art visual perception solutions, like building maps of the environment or accurately tracking movement over time. However, it also adds weight and cost.
At the MAVLab, the aim is to make light-weight and cheap autonomous racing drones. These could be used by many drone racing enthusiasts to train with or fly against. If they become small enough, they could even be used for racing at home.
The main innovation underlying the world’s smallest drone is the creation of extremely efficient and yet still robust algorithms. “The wireless images in human drone racing can be very noisy and sometimes not even arrive at all,” says MAVLab founder Christophe De Wagter. “So, pilots rely heavily on their predictions of how the drone is going to move when they move the sticks on their remote control.”
Although the images of an autonomous drone do not have to be transmitted through the air, their interpretation by small drones can sometimes be completely off. The drone can miss a gate or evaluate its position relative to the gate wrongly. For this reason, a prediction model is central to the approach. Since the drone has very little processing, the model only captures the essentials, such as thrust and drag forces on the frame.
“When scaling down the drone and sensors, the measurements deteriorate in quality, from the camera to the accelerometers,” explains Shuo Li, PhD student at the MAVLab on the topic of autonomous drone racing. “Hence, the typical approach of integrating the accelerations measured by the accelerometers is hopeless. Instead, we’ve only used the estimated drone attitude in our predictive model. We correct the drift of this model over time by relying on the vision measurements.” A new robust state estimation filter was used to combine the noisy vision measurements in the best way with the model predictions.
The drone used the newly developed algorithms to race along a 4-gate race track in TU Delft’s Cyberzoo. It can fly multiple laps at a competitive average speed of 2 m/s. Thanks to the central role of gate detections in its algorithms, it can cope with displacements of the gates.
“We’re currently still far from the speeds obtained by expert human drone racers. The next step will require even better predictive control, state estimation and computer vision,” says De Wagter. “Efficient algorithms to achieve these capabilities will be essential, as they’ll allow the drone to sense and react quickly. Moreover, small drones can choose their trajectory more freely, as the racing gates are relatively larger for them.”
Autonomous racing drones are useful in other domains as well. “For typical drones with four rotors, flying faster also simply means that they’re able to cover more area. For some applications, such as search and rescue or package delivery, being quicker will be hugely beneficial,” adds Guido de Croon, scientific leader of the MAVLab. “Our focus on light-weight and cheap solutions means that such fast flight capabilities will be available to a large variety of drones.”
Main photo: Yingfu Xu, TU Delft