Site icon Wonderful Engineering

New Transformer-Style Humanoid Robot Can Launch A Shapeshifting Drone Off Its Back

Caltech engineers have unveiled a robot system that looks like it stepped out of a sci-fi movie: a humanoid robot carrying a transforming drone on its back that can launch, drive, and fly depending on what the mission demands, as reported by LiveScience. The setup pairs a Unitree G1 humanoid with a drone called M4, creating a multimodal robot capable of tackling environments no single machine could handle alone.

The project is the result of a three-year collaboration between Caltech’s Center for Autonomous Systems and Technologies and the Technology Innovation Institute in Abu Dhabi. Inspired less by Optimus Prime and more by Soundwave, the humanoid serves as a mobile deployment platform while the drone handles fast, adaptable movement.

The humanoid can walk, climb stairs, and navigate toward the drone’s destination, though it moves more slowly than other advanced humanoids. The real star is M4, which can transform its body depending on the terrain. It can roll on four wheels, flip into a meerkat-style stance on two wheels, walk by rotating its wheels like feet, and even use two rotors to help climb steep slopes. When flight is needed, all four wheels fold upward as propellers take over.

Aaron Ames, director of CAST, says the goal is to combine robotic movement types that typically exist separately. Robots can walk, robots can drive, and robots can fly, but merging all three into one system unlocks entirely new capabilities. The challenge, he explains, was integrating two different machines so they act as a single coordinated system while still offering unique functions.

To launch the drone, the humanoid bends forward, allowing M4 to lift off directly from its back. From there, M4 can scout ahead, enter difficult terrain, or navigate tight environments the humanoid cannot safely reach.

Researchers say the next step is reliability. They want future robots to operate safely around people, withstand real-world environments, and make decisions with secure, dependable control systems. This collaboration is part of a broader push to solve those long term autonomy challenges and move robotic systems closer to everyday use.

Exit mobile version