New low-cost technology to avoid collisions between drones

Most drones we know in everyday life are controlled remotely by someone, even if they are out of sight of the operator. For a drone to be fully autonomous, it must be able to make flight decisions on its own and without human intervention, that is, decide for itself how to avoid collisions, how to maintain a course in the face of wind gusts, control flight speed, avoid buildings, trees, etc.

The Computational Intelligence Group (GIC) of the University of the Basque Country (UPV/EHU) has designed a new low-cost technology that can be reproduced for all types of aerial robots.

This new low-cost autonomous navigation technology was developed by Julián Estévez and his collaborators.

The new technology prevents two or more drones crossing paths in the air from colliding with each other. And it uses only onboard sensors and cameras to do its job.

In the tests carried out so far, the results have been very positive and encouraging.

The authors of the new study have verified that, “despite the low cost of the technology, the solution we have developed has been successfully validated in commercial drones. Using simple, low-cost equipment and an algorithm based on artificial vision, based on colour identification, we have developed a robust technology to successfully avoid collisions between drones that can be easily extrapolated to most commercial and research aerial robots; in addition, we provide the complete software code of the solution,” says Julián Estévez, a researcher at the University of the Basque Country.

“This work is a small step towards fully autonomous navigation – without any human intervention – and allowing drones to decide which manoeuvre to perform, which direction to take and thus avoid collisions between themselves or with other air obstacles. If we accept that, in the future, our airspace will be much more populated by commercial services carried out by these drones, our work is a small contribution in this direction,” says Julián Estévez.

The author of the study explains that “our proposal for avoiding collisions does not require the drones to exchange information between them; instead, they rely solely on onboard sensors and cameras.” “We obtain the signal from the camera on board the drones and, by processing the images, we adjust the reactions of the robots so that they fly smoothly and accurately,” adds Estévez.

In the experiments they have tried to imitate realistic conditions of drones, that is, scenarios that can occur in a typical urban area under uncontrolled lighting conditions, drones flying in different directions, etc., so their contributions are focused on applications in the real world, despite the initial work in the laboratory.

The UPV/EHU Computational Intelligence Group has developed a low-cost technology to prevent two or more drones from colliding with each other. (Photo: UPV/EHU)

Color-based algorithms

“We have equipped each drone with a red card that allows the software algorithm to detect the presence of an approaching drone and measure its proximity,” explains Julián Estévez. “Our proposal,” the researcher continues, “is very simple: each drone is equipped with an onboard camera, whose screen is divided into two halves (left and right). This camera is constantly looking for the red color of the cards we have indicated above. Through simple image processing, we can know what percentage of the camera is occupied by the red color, and whether most of this red region is on the left or right of the screen. If most of the red area is on the left side of the screen, the drone will fly to the right to avoid the collision. If the red area is on the right, it will fly to the left. And this happens with all drones in the air.”

Furthermore, “when the percentage of red on the screen increases, this will mean that the drones are approaching head-on. So when a threshold is exceeded, the robot will know that it must perform the evasive maneuver. All this happens autonomously, without intervention from the human operator. It is a simple way of avoiding collisions, capable of being carried out using sensors and low-cost equipment,” emphasizes Julián Estévez. Something similar to what happens when a person is walking down the street and sees someone approaching from the left, in this case the person tries to move to the right to avoid colliding.

The study is titled “A low-cost vision system for online reciprocal collision avoidance with UAVs”. It has been published in the academic journal Aerospace Science and Technology. (Source: UPV/EHU)