New research published in Science Robotics showcases upgrades that allow the Robobee to land with the grace of a crane fly (image credit Science Robotics)

A miniature marvel of bioinspired engineering, the Harvard RoboBee, has long demonstrated its ability to zip, dive, and hover like a real insect. But even the most sophisticated aerial robots face one enduring challenge: how to land safely.

Hyun’s work included conducting precise landing tests on natural and rigid surfaces and enhancing the controller—the robot’s “brain”—to recognize and adapt to changing airflow conditions during descent. The result: a dramatically improved landing process that protects the robot’s fragile piezoelectric actuators, the ultra-light “muscles” that power its flight.

RoboBee’s landing instability was caused by the robot’s minuscule size (just a tenth of a gram) and the turbulence created by its flapping wings near the ground (photo credit Harvard/Christian Chan)

The research team, led by Harvard’s Robert Wood, the Harry Lewis and Marlyn McGrath Professor in the John A. Paulson School of Engineering and Applied Sciences (SEAS), drew inspiration from the crane fly—an insect known for its slow, stable landings and long, jointed legs. The robot’s new appendages mimic this structure, helping absorb impact and stabilize the landing. The legs were crafted using custom fabrication techniques pioneered at Harvard’s Microrobotics Lab.

“This project perfectly illustrates how electrical and computer engineering research at Purdue connects deeply with advances in robotics,” Hyun said. “We’re building more than robots—we’re building platforms for testing biological theories and advancing autonomy.”

With tethered flight still a limitation, the team’s longer-term goal is full autonomy: onboard power, sensors, and computing capabilities that would allow RoboBee to fly and land independently.

“In the interim we have been working through challenges for electrical and mechanical components using tethered devices,” said Wood. “The safety tethers were, unsurprisingly, getting in the way of our experiments, and so safe landing is one critical step to remove those tethers.”

Hyun’s ongoing research at Purdue is aimed squarely at those challenges, generalizing the controllability of multi-scaled flapping-based mechanisms to even bigger bird scales. “Seemingly simple periodic wing flapping motion of birds and insects provoked many scientists and engineers to wonder that the flapping motion is purely an open-loop process. However, the hovering or the navigation via wing flaps is extremely unstable, requiring continuous effort to correct its flapping pattern to stay afloat. This is why I like studying animal locomotion and designing flexible decision-making to enhance controllability. Nature provides a beautiful existence proof of how locomotive systems can be efficiently controlled, and we are trying to bring mathematical rigor to advance the synthetic systems on fast-adaptive decision-making. ”

The potential applications for such insect-scale flying robots are vast, including environmental monitoring, disaster response, and artificial pollination. With continued contributions from Purdue ECE, RoboBee and future microrobots are buzzing closer to reality.

This work was supported by the National Science Foundation Graduate Research Fellowship Program under Grant No. DGE 2140743.

Source: Sticking the landing: Insect-inspired strategies for safely landing flapping-wing aerial microrobots