Researchers at the University of Houston have developed a new safety system that can help drones avoid crashes while they are flying.
The technology constantly watches the drone’s movement and can step in immediately if it detects that the drone is heading toward danger.
The new system was created by Marzia Cescon, an assistant professor of Mechanical and Aerospace Engineering at the University of Houston’s Cullen College of Engineering.
Her research was published in the Journal of Dynamic Systems, Measurement, and Control.
Drones are now used in many industries. They inspect bridges, power lines, and buildings, take aerial photographs and videos, deliver packages, and assist with emergency response.
One of the most common types is the quadrotor, also called a quadcopter. It has four spinning rotors that allow it to fly in different directions, hover in one place, and move through narrow spaces with great precision.
Although modern drones are highly capable, they are not immune to unexpected events.
Strong wind gusts, sudden obstacles, or errors in navigation software can push a drone off its planned flight path. When this happens, the drone could collide with buildings, trees, power lines, or other objects, putting both the drone and people nearby at risk.
To improve safety, Cescon developed what she calls a “safety supervisor.” This is a small software module that runs on the drone while it is flying. Instead of replacing the drone’s normal flight controls, it works alongside them, continuously checking whether the drone is operating safely.
The system constantly monitors important information, including the drone’s position and how much it is tilting. By analyzing this data in real time, it predicts whether the drone is getting too close to a dangerous area before an accident happens.
Cescon compares the system to an invisible safety fence. As long as the drone stays inside the safe area, the supervisor simply watches without interfering. But if it predicts that the drone is about to cross the invisible boundary and risk crashing, it immediately steps in and gently changes the drone’s flight path to keep it safe.
The technology is based on a mathematical method called a Control Barrier Function. Although the mathematics behind it is complex, its purpose is simple: to continuously calculate whether the drone is approaching danger and, if necessary, briefly take control to prevent an accident.
The safety supervisor was developed and tested in the University of Houston’s Advanced Learning, Artificial Intelligence and Control Laboratory. The research also demonstrated that the system works well with the flight controllers already used in many drones, making it practical for real-world applications.
As drones become more common in transportation, inspections, deliveries, agriculture, and emergency services, safety will become increasingly important.
A system that can monitor every flight, react instantly to unexpected situations, and help prevent crashes could make drone operations more reliable and give pilots and operators greater confidence.
The researchers believe this technology is an important step toward safer autonomous drones that can operate in complex environments with less risk.
