Engineers at Northwestern University have developed a drone that uses rapid rotation and motion blur to make its structure less visible during flight. The prototype, called Phantom Twist, spins its body as many as 25 times per second. At that speed, the drone’s individual components become difficult for the human eye to distinguish, causing the aircraft to appear as a faint, semi-transparent blur rather than a fixed shape.
The researchers said the approach could reduce the visual disruption caused by drones used for wildlife monitoring, environmental surveys and infrastructure inspections. Drones operating in those settings can affect the behaviour of people and animals when they are noticed.
Previous efforts to make drones less visible have generally relied on camouflage, transparent materials or optical systems designed to redirect light. The Northwestern team instead designed the aircraft around the way human vision processes rapidly moving objects.
“Most efforts to hide drones focus on making them look like their surroundings,” said Michael Rubenstein, an associate professor of computer science and mechanical engineering at Northwestern’s McCormick School of Engineering. “Instead, we asked whether we could design the drone itself around the way humans perceive motion.”
Unlike a conventional quadcopter, which has a stationary body supported by four rotating propellers, Phantom Twist uses one motor and one propeller. The propeller rotates in one direction while the rest of the aircraft rotates in the opposite direction, leaving no major stationary section visible during flight.
The research team used computational modelling to generate about 20,000 configurations capable of stable flight. Optimization algorithms then adjusted the positions of the motor, propeller, circuit board, batteries and counterweight to reduce the drone’s visibility from different viewing angles.
The researchers simulated the designs while spinning and placed the resulting images over 100 real-world backgrounds. A perception model intended to approximate human vision assigned each configuration a visibility score. The team selected the 500 lowest-scoring designs and conducted further optimization before building the prototype.
The final design distributes its components across different heights and angles, with gaps intended to prevent the parts from visually overlapping as the aircraft rotates. When blurred by the motion, the opaque components are visually averaged with the background.
“The human eye takes time to accumulate signals, roughly analogous to the exposure time of a camera,” said Emma Alexander, an assistant professor of computer science at Northwestern. “When an object spins quickly, we perceive it as blurring out and losing distinct features.”
According to the researchers’ perception metric, the optimized design was about 10 times less visually perceptible than a conventional quadcopter. The comparison was based on the computational visibility model rather than a claim that the drone is completely invisible.
The prototype retains several limitations. Its propeller produces audible noise, while its wires and support rods can still be seen. The researchers plan to examine the use of more transparent materials and quieter propulsion systems in future versions.
The project also involved Northwestern researchers Sam Kriegman, David Matthews, Jingxian Wang and Chen Yu.
