top of page
Image: MIT. The 1.29 gram flapping wing robot landing on a sunflower, outside the lab's motion capture rig

Tiny Flying Robot Packs Own Brain

Writer: Marcus Tan
Marcus Tan
11 minutes ago
3 min read

MIT researchers have flown a flapping wing robot weighing 1.29 grams that carried its own sensors and computation, then landed precisely on a sunflower. The machine is roughly the size and shape of a microcassette and weighs less than a paperclip. It still draws power through a tether.


Image: MIT. The 1.29 gram flapping wing robot landing on a sunflower, outside the lab's motion capture rig
Image: MIT. The 1.29 gram flapping wing robot landing on a sunflower, outside the lab's motion capture rig

The work, reported by MIT News and published in Science Robotics, was led by Kevin Chen, an associate professor at MIT and head of its Soft and Micro Robotics Laboratory. He calls it the first demonstration of an aerial microrobot that is sensory and compute autonomous. Earlier flyers could match a bumblebee, but only inside an arena where offboard cameras tracked them.


Lifting the electronics needed more thrust than the earlier design produced. The team grew the soft artificial muscles that drive the wings by 50 percent and raised net lift by 46 percent. And that pushed the lift to weight ratio to 1.9, enough to haul a 244 milligram electronics package mounted on a fiberglass plate.


The package holds three sensors and the chip that flies the robot. An inertial measurement unit tracks attitude, an optical flow sensor tracks lateral speed and position, and a time of flight sensor measures height. A microcontroller runs the estimation and control loop on the vehicle.


Without the cameras the flyer has no absolute position and drifts. The team simplified the maths so position updates arrive faster, and found a rough estimate at a high rate beat a precise one that lagged. Yi-Hsuan Hsiao, one of the paper's lead authors, says the robot flies with accuracy comparable to offboard cameras.


Positional errors stayed within a few centimetres. The electronics account for about 6 percent of power use, with the muscles and wings taking the rest. But the chip must handle perception and control while riding an airframe that cannot carry much, which Jonathan How, a professor of aeronautics and astronautics, describes as a problem where every part has to be thought through.


For the pollination test an operator sent position setpoints from a joystick, and the robot flew around several plants before settling on the sunflower. The paper calls that a hierarchical controller, with the human picking destinations and the onboard loop holding flight.


The team is now working on an onboard camera for object recognition and obstacle avoidance, plus power electronics and small batteries to cut the tether. The work was funded by the US National Science Foundation, the Office of Naval Research, the Air Force Office of Scientific Research, MathWorks, and a Zakhartchenko Fellowship.


So the robot senses and computes in the air, lands on a flower away from the camera rig, and still takes electricity down a wire. Power autonomy is the part the team is still working on.


  • A 1.29 gram flapping wing robot flew with onboard sensing and computation, landing on a sunflower without motion capture cameras.

  • Enlarged artificial muscles lifted net thrust by 46 percent, enough to carry a 244 milligram sensor and microcontroller package.

  • Positional errors stayed within a few centimetres, with the electronics drawing about 6 percent of total power.

  • The flyer remains tethered to an external power supply, and the team is working on batteries and an onboard camera.

  • The research appears in Science Robotics.


 
 

Technology increasingly permeates every facet of our lives, making informed decision making an essential pursuit. We bridge this gap by combining the precision of AI with the irreplaceable discernment of human expertise. Our team produces rigorous product reviews that offer unique insights, honest critiques, and trustworthy recommendations. We also leverage AI to synthesise complex news from reliable sources into clear, actionable updates, ensuring that every story is carefully fact checked by our editorial staff before publication. Accuracy remains our priority. Should you identify any discrepancies, please contact us at editorial@tech360.tv. Your feedback is a vital part of our process in maintaining the high standards our readers deserve.

Tech360tv is Singapore's Tech News and Gadget Reviews platform. Join us for our in depth PC reviews, Smartphone reviews, Audio reviews, Camera reviews and other gadget reviews.

  • YouTube
  • Facebook
  • TikTok
  • Instagram
  • Twitter
  • LinkedIn

© 2021 tech360.tv. All rights reserved.

bottom of page