RoboBee engineers continue to innovate and have now developed technology for better legs for the microrobot to land.
By Reshmi Thampy
Microrobots inspired by flying insects don’t always stick the landing—until now.
Inspired by real bees, RoboBee is a microrobot designed by researchers at Harvard’s John A. Paulson School of Engineering and Applied Sciences. Weighing a tenth of a gram with a wingspan of 3 cm, RoboBee made headlines with its debut in 2013 and with each iteration and upgrade. This year, the RoboBee team successfully eliminated a threat to its delicate actuators (the part that moves the wings) with a new improvement to its structure—long articulated legs that can cushion its fall during landing.
Meet RoboBee
Unlike rigid robots, RoboBee has a soft, lightweight body driven by piezoelectric actuators—thin layered structures that bend when voltage is applied. These “artificial muscles” connect to a carbon frame. Up to 20 thin ceramic layers are used because a larger surface area means less voltage is needed for activation. When activated, the electrodes squeeze the ceramic layers between them. This mechanical strain is used to power the attached wings.
These wings are around 20 micrometers thick—roughly a quarter the width of a human hair—and flap at 170 Hz (170 times a second) to launch the microrobot into the air and hover. But the downward draft generates eddies as the bee nears the ground. This ground effect is significant since the RoboBee is lightweight, and the impact during landing could mean a death sentence for the tiny bot.
Cross your fingers and hope to land
In previous versions, the RoboBee made the landing by hovering near the ground as close as possible, and powering off—hoping that it landed upright. However, a working bee should be able to transition between different environments (air, water, land, and leaf to name a few), and a hit-or-miss landing won’t do. To solve this problem, the team turned to the designs of the oldest engineer—nature.
Winging it like a bug—using biomimicry
Insects show a diverse landing behavior—from those that graze or hover to slowly decelerate before landing to the energetic ones that accelerate and land aggressively. Some insects like mosquitoes have long forelegs with appendages that can disperse the impact force, sometimes causing it to bounce.
RoboBee’s initial version had four rigid carbon fiber sticks of arbitrary length and placement for the sole purpose of landing the bee upright. The new legs were inspired by a crane fly—they have long legs like a mosquito, but their wingspan and body length are similar to our RoboBee. Crane flies have to land frequently as they are weak fliers, and their long legs damp the impact energy effectively during landing—exactly what we want in our microrobot.
Researchers examined the leg proportions of 27 crane fly species to arrive at the ideal joint placement and segment lengths for the RoboBee. The team tested 12 different leg designs, comparing factors like coefficient of restitution (CoR)—a measure of energy dissipation during landing, horizontal drift, and landing success. The legs are built with viscoelastic joints made from Kapton films (a flexible polyimide acting as the spring element) sandwiched between two layers of carbon fiber composite, laminated by two layers of viscoelastic thermoplastic elastomer (damping element) on the outside. The winning design was the one that showed the least CoR (maximum energy dissipation), allowing the bot to maintain contact with the landing surface without bouncing.
Bee Brain gets a software update
New legs are not the only upgrade for the RoboBee to land right. Insects use optic flow—the speed at which objects get bigger in their eyes is used to safely decelerate before landing. Engineers adopted the same technique in the RoboBee as they developed a time-to-contact model. This allows the bot to power off its wings before impact.
RELATED: Running Robots May Be Inspired by Birds
Bee-yond expectations
RoboBee has gained the title of smallest insectoid robot capable of controlled flight. But engineers have come up with more upgrades and designs, including a bee with cross wings to make it capable of spinning or yaw, called the X-wing bee. Another, called Bee+, has four wings instead of two and requires less power. They also have one that can land on water; another that can adhere to surfaces using static electricity; and a solar-powered X-wing bee with four wings that can fly untethered for short periods under intense lighting. But RoboBee is still dependent on a tether that supplies its power and flight commands. To become a working bee, it has to get rid of this umbilical cord. Ongoing research has a goal of making this possible by miniaturizing components called the triple Holy Grail—power supply, control system, and sensors.
To bee or not to bee—potential applications of the RoboBee
Electrical engineer Gu-Yeon Wei came up with the idea of RoboBee after he came across an article on colony collapse disorder afflicting honeybees. (This is when worker bees disappear, leaving the queen and a few nurse bees and immature bees to fend on their own.) He teamed up with Robert Wood, who had just successfully created a robotic fly. The idea was to create a swarm of robotic bees that could pollinate flowers as a colony of bees would do.
Greenhouses introduce bees to pollinate, but their efficiency is low. Bees get disoriented, hit the walls of the greenhouse, and die. Swarms of RoboBees pollinating vertical farms and greenhouses is a possible application for RoboBees. However, environmentalists and bee conservation groups warn that it is not economically feasible and could lead to food insecurity.
Using RoboBees for surveillance is another possible application that seems straight out of a dystopian novel.
While the idea of retiring bees needs debate, a more ethical application is a swarm of RoboBees for search and rescue applications. It could be well suited to find potential survivors in the aftermath of a disaster, enabled by its small size and hive brain.
No matter what RoboBees wanna-bee in the future, they are worth the buzz.
This study was published in the peer-reviewed journal Science Robotics.
Which insectoid would you like to see realized? The author would love a pet spider-robo that can spin webs like a mini 3D printer with legs.
References
Hyun, N. P., Chan, C. M., Hernandez, A. M., & Wood, R. J. (2025). Sticking the landing: Insect-inspired strategies for safely landing flapping-wing aerial microrobots. Science Robotics, 10 (101), eadq3059. https://doi.org/10.1126/scirobotics.adq3059
Shaw, J. (2024, August 13). Building RoboBees: How Harvard engineers are revolutionizing micro-robotics. Harvard Magazine. https://www.harvardmagazine.com/2017/10/harvard-robot-bees-future-robotic-engineering
Featured image of RoboBees landing on a leaf, credit: Harvard Microrobotics Laboratory. Illustration credit: Reshmi Thampy.
About the Author
Reshmi Thampy is a postgraduate in structural and construction engineering. She loves all things green and would have been a farmer in a parallel universe. In this one, she is an avid reader and an occasional writer. She loves reading about new advances in science and technology and views the field of science journalism as an exciting one where she can share her curiosity with others. You can reach her at reshmithampy@proton.me.
