A new generation of robotics is split between two extremes: making machines capable of human-like combat and making drones that are virtually invisible to the naked eye. While these technologies sound like military science fiction, they are grounded in very different engineering challenges.
Startups like Foundation Future Industries are working to build humanoid robots for the military, aiming to field machines that can navigate rubble or enter buildings. Meanwhile, researchers at Northwestern University have developed a drone called Phantom Twist, which relies on high-speed spinning to become a transparent blur to human observers. While the humanoid tech remains in early, experimental stages, the drone's invisibility is an established result of a process called computational design.
The math behind the invisible drone
To make a drone invisible, engineers use a process where a computer runs through thousands of potential layouts to find one that tricks the human eye. We have a limited ability to track fast motion; if an object spins at 15 to 25 times per second, our eyes fail to focus on it, instead seeing a transparent smear. Computers optimize the drone's physical parts—batteries, motors, and carbon fiber rods—to be as spread out as possible when spinning. By minimizing the overlap of these parts, the drone effectively lets the background show through. This is far more effective than trying to camouflage the color or shape of the craft.
In contrast, humanoid robots like those being developed for combat rely on AI algorithms for movement and balance. Training these machines involves teaching them to navigate complex, changing environments by moving limbs in real-time. The biggest hurdles here are not just software, but physical limits. Manipulating objects, like picking up a rifle or climbing stairs, requires extreme precision and sensitivity that current machines rarely achieve outside of controlled lab settings. While companies claim these machines are ready for deployment, experts note that turning a lab-proven robot into a soldier capable of traversing real-world debris remains a challenge that likely spans a decade or more.
These developments highlight a growing gap between the promise of artificial intelligence and the reality of physical machines. We are getting better at training robots to solve abstract movement puzzles or design efficient shapes. However, building a machine that can reliably operate in messy, unpredictable places remains a major hurdle. While military interest attracts investment for armed humanoids, the most immediate impact of this research might be in surveillance. As drones become quieter and harder to detect, the simple act of flying a craft becomes a tool for new levels of observation, carrying with it significant questions about privacy and the nature of autonomous systems in our daily lives.
Liked this one? The next lands at breakfast.
Every story in tomorrow's AI news, rebuilt in plain English — five minutes, sources linked, free forever.
By joining you agree to receive Article's daily newsletter — unsubscribe in one click. Privacy