Dutch engineers develop robots, flying as agile as flies

Dutch engineers have developed a lightweight flying-wing robot inspired by insects that can hover, move in different directions and perform highly agile manoeuvres, potentially opening the door to a new generation of drones for surveillance, exploration and other applications.

The robot, developed by researchers at Delft University of Technology, uses rapidly flapping wings rather than conventional propellers to generate lift and control its movement. The research was published in the journal Science.

The 29-gram robot has a wingspan of 33 cm and beats its wings around 17 times per second. Like flying insects, it can make small changes to its wing movements to control its direction and remain stable in the air.

According to Matej Karasek, the study's first author and the robot's main designer, the machine can reach a top speed of 25 kmph and perform demanding aerial manoeuvres, including 360-degree flips that resemble loops and barrel rolls.

The robot's low weight also gives it an advantage in terms of energy efficiency. Researchers said it can hover for around five minutes on a fully charged battery or cover a distance of more than 1 km.

What makes the technology particularly interesting is that its movements closely resemble those of fruit flies. By studying the insects' rapid and precise changes in wing motion, the researchers were able to reproduce some of their remarkable flying abilities in a robotic system.

The robot also demonstrated how fruit flies adjust their turning angle to improve their chances of escaping predators. This could help scientists better understand insect flight while providing engineers with new ideas for developing small, highly manoeuvrable aerial robots.

Unlike conventional drones, which generally rely on rotating propellers and can struggle with certain rapid movements, flapping-wing robots could eventually be useful in environments where agility, compact size and precise movement are more important than carrying heavy loads.

The research therefore represents not only a step forward in drone design but also an example of how studying nature can inspire solutions for advanced robotics and autonomous flight.