New algorithm lets autonomous drones work together to transport heavy, changing payloads

Scientists at TU Delft, the Netherlands, have developed a new algorithm that allows multiple autonomous drones to work together to control and transport heavy payloads, even in windy conditions. Drones are ideal for reaching and maintaining hard-to-reach infrastructure, like offshore wind turbines. With often harsh weather, limited payload capacity and unpredictable contact with the environment, it is difficult for current drones to operate safely and effectively.

“A single drone can only carry a very limited load,” explains Sihao Sun, robotics researcher at TU Delft. “This makes it hard to use drones for tasks like delivering heavy building materials to remote areas, transporting large amounts of crops in mountainous regions, or assisting in rescue missions.”

To overcome these limitations, the TU Delft team designed a system where multiple drones are connected to a payload via cables, thereby carrying much heavier loads. By adjusting their positions in real time, the drones can not only lift and transport the heavy object but also control its orientation, which is crucial for precise placement in complex environments. The study has been published in Science Robotics.

Fast coordination

“The real challenge is the coordination,” says Sun. “When drones are physically connected, they have to respond to each other and to external disturbances like sudden movements of the payload in rapid motions. Traditional control algorithms are simply too slow and rigid for that.”

The new algorithm developed by the team is fast, flexible, and robust. It adapts to changing payloads and compensates for external forces without requiring sensors on the payload itself, which is an important advantage in real-world scenarios.

Basketball obstacle course in the lab

“We built our own quadrotors and tested them in a controlled lab environment,” Sun says. “We used up to four drones at once, added obstacles, simulated wind with a fan, and even used a moving payload like a basketball to test dynamic responses.”

The system passed all tests. And because the drones are autonomous, they only need to be given a destination. They navigate independently, adjusting for obstacles and disturbances along the way. “You just tell them where to go, and they figure out the rest,” Sun adds.

Preparing for real-world applications

Currently, the system uses external motion capture cameras for indoor testing, and is therefore not useful in outdoor environments yet. The team hopes to prepare the technology for real-world deployment in the future, with potential applications in search and rescue, agriculture, and remote construction.