Humanoid robots have mostly been stuck with walking, grabbing, or at best, tripping through obstacle courses. But the Italian Institute of Technology (IIT) in Genoa smashed that limit with iRonCub3, the world’s first jet-powered humanoid robot to nail controlled flight.
Hovering 50 centimeters off the ground, this 70-kilogram engineering wonder mixes a human-like shape with sky-high skills, hitting a milestone straight out of a sci-fi flick. iRonCub3 is the third take on IIT’s iCub platform, a kid-sized humanoid built for teleoperation. This time, it’s packing four microjet turbines—two on the arms, two on a backpack-style jetpack—kicking out over 1,000 newtons of thrust.
DJI Neo, Mini Drone with 4K UHD Camera for Adults, 135g Self Flying Drone that Follows You, Palm Takeoff,...
- Lightweight and Regulation Friendly - At just 135g, this drone with camera for adults 4K may be even lighter than your phone and does not require FAA...
- Palm Takeoff & Landing, Go Controller-Free [1] - Neo takes off from your hand with just a push of a button. The safe and easy operation of this drone...
- Subject Tracking & QuickShots - Effortlessly capture stunning vlogs as DJI Neo smartly follows you. Getting professional footage has never been easier...
Daniele Pucci, who leads IIT’s Artificial and Mechanical Intelligence (AMI) Lab, lays out the challenge: “This research is radically different from traditional humanoid robotics and forced us to make a substantial leap forward with respect to the state of the art. Thermodynamics plays a pivotal role—the emission gases from the turbines reach 700°C and flow at nearly the speed of sound.” To handle that fiery ordeal, the team gave iRonCub3 a titanium spine and heat-shield covers, letting it shrug off exhaust temps up to 800°C without turning into a molten mess.
Keeping a humanoid steady in the air is no cakewalk. Drones lean on symmetry and ease, but iRonCub3’s long frame and moving limbs shift its balance with every move. Antonello Paolino, the study’s lead author and a PhD student at IIT and Naples University, says, “Our models include neural networks trained on simulated and experimental data and are integrated into the robot’s control architecture to guarantee stable flight.”
The crew cooked up cutting-edge control algorithms with constrained Quadratic Programming, backed by Lyapunov analysis, to juggle slow joint motors with the zippy jet turbines. Wind tunnel runs at the Polytechnic of Milan and AI-powered aero models from Stanford University fine-tuned its hover skills, even in windy bursts.
Building iRonCub3 meant a total overhaul of the iCub’s blueprint. The team used PTC Creo to whip up a CAD model, constantly tweaked with real-world test data. A semi-auto URDF process turned that into kinematic and dynamic models for planning and sims. At 70 kilograms, it’s a beefier leap from past versions, yet its jets lift it off with pinpoint accuracy. Early trials in IIT’s indoor lab worked like a charm, with iRonCub3 holding steady during short hovers. Next up, they’re gearing to stretch limits at a special flight zone at Genoa Airport, teaming with Aeroporto di Genova to roll out safety-checked, bolder moves.
[Source]
