Engineers at Northwestern University have developed an unconventional drone that becomes remarkably difficult for the human eye to see while flying. Called Phantom Twist, the experimental aircraft relies on rapid spinning and artificial intelligence-driven design rather than camouflage or transparent materials. The project demonstrates a novel approach to low-visibility flight by taking advantage of the limitations of human visual perception instead of attempting to hide the drone physically, tells IEEE Spectrum.
Unlike a conventional quadcopter, Phantom Twist uses a single motor and propeller while its entire body rotates between 15 and 25 times per second. This rapid motion creates a persistence-of-vision effect, causing the drone to appear as a faint, transparent blur rather than a clearly defined object. Because much of the drone’s structure consists of open space connected by thin carbon-fiber rods, the spinning motion allows the background to remain visible through the aircraft, making it significantly harder to detect.
The researchers used computational design to optimize the drone’s geometry. An iterative algorithm evaluated approximately 20,000 feasible configurations, balancing flight stability with minimal visual detectability. The optimization relied on a perceptual image similarity metric that compared background scenes with simulated drone images, selecting layouts that produced the smallest visual difference. The resulting design is reported to be more than ten times less visible than a similarly sized conventional quadcopter.
Despite its visual advantages, Phantom Twist remains an experimental platform. It currently depends on an optical tracking system for controlled flight and still generates the characteristic noise associated with drone propulsion. The researchers believe future versions could reduce both visual and acoustic signatures while incorporating onboard sensing. They also suggest that a camera mounted on the spinning body could capture panoramic imagery for navigation and environmental awareness.
Beyond potential surveillance applications, the team envisions quieter, less visually intrusive drones for wildlife observation and environmental monitoring, where minimizing disturbance is essential. The work highlights how AI-guided engineering and human-centered design principles can create unconventional solutions by optimizing not only mechanical performance but also the way machines are perceived by people.