Close Menu
    Facebook X (Twitter) Instagram
    SciTechDaily
    • Biology
    • Chemistry
    • Earth
    • Health
    • Physics
    • Science
    • Space
    • Technology
    Facebook X (Twitter) Pinterest YouTube RSS
    SciTechDaily
    Home»Technology»MIT’s Tiny Flying Robot Just Learned To Move Like a Real Insect
    Technology

    MIT’s Tiny Flying Robot Just Learned To Move Like a Real Insect

    By Adam Zewe, Massachusetts Institute of TechnologyJuly 21, 2026No Comments7 Mins Read
    Facebook Twitter Pinterest Telegram LinkedIn WhatsApp Email Reddit
    Share
    Facebook Twitter LinkedIn Pinterest Telegram Email Reddit
    Microrobot on Hand
    Chen’s group has been building robotic insects for more than five years. Credit: Courtesy of the Soft and Micro Robotics Laboratory

    Moving with the speed and agility of an insect, the tiny robot could someday assist in search-and-rescue missions.

    After an earthquake, survivors can become trapped beneath piles of unstable debris where conventional rescue robots are too large to enter. Tiny flying machines modeled on insects could one day move through these narrow gaps while avoiding fixed barriers and falling rubble.

    Until recently, however, aerial microrobots could only travel slowly along simple, smooth paths, falling far short of the rapid and agile flight seen in insects — until now.

    AI unlocks insectlike aerial agility

    MIT researchers have demonstrated aerial microrobots that can match the speed and agility of their biological counterparts. They developed a new AI-based controller that allowed the robotic insect to perform demanding flight maneuvers, including repeated body flips.

    The two-part control system combines strong performance with computational efficiency. Compared with the researchers’ best earlier demonstrations, it increased the robot’s speed by about 450 percent and its acceleration by roughly 250 percent.

    Flying Microrobot Performing a Flip
    A time-lapse photo shows a flying microrobot performing a flip. Credit: Courtesy of the Soft and Micro Robotics Laboratory

    The microrobot completed 10 consecutive somersaults in 11 seconds while maintaining control even as wind disturbances pushed against it.

    Flight performance approaches living insects

    “We want to be able to use these robots in scenarios that more traditional quadcopter robots would have trouble flying into, but that insects could navigate. Now, with our bioinspired control framework, the flight performance of our robot is comparable to insects in terms of speed, acceleration, and the pitching angle. This is quite an exciting step toward that future goal,” says Kevin Chen, an associate professor in the Department of Electrical Engineering and Computer Science (EECS), head of the Soft and Micro Robotics Laboratory within the Research Laboratory of Electronics (RLE), and co-senior author of a paper on the robot.

    Chen is joined on the paper by co-lead authors Yi-Hsuan Hsiao, an EECS MIT graduate student; Andrea Tagliabue PhD ’24; and Owen Matteson, a graduate student in the Department of Aeronautics and Astronautics (AeroAstro); as well as EECS graduate student Suhan Kim; Tong Zhao MEng ’23; and co-senior author Jonathan P. How, the Ford Professor of Engineering in the Department of Aeronautics and Astronautics and a principal investigator in the Laboratory for Information and Decision Systems (LIDS). The research was published in Science Advances.

    A robotic insect no larger than a microcassette and weighing less than a paperclip can now perform rapid, demanding flight maneuvers. The achievement builds on more than five years of work by Chen’s group to improve both the physical design and control of tiny flying robots.

    Insect Scale Robot Flight Trajectory
    “Now, with our bio-inspired control framework, the flight performance of our robot is comparable to insects in terms of speed, acceleration, and the pitching angle,” says Kevin Chen. Credit: Courtesy of the Soft and Micro Robotics Laboratory

    Their latest model is more durable than earlier versions and has larger flapping wings for greater agility. Soft artificial muscles power those wings, driving them back and forth at extremely high speeds.

    The hardware had advanced, but the robot’s controller — the “brain” that tracks its position and directs its flight — remained a major limitation. Because a person had to tune it by hand, the system could not fully exploit the robot’s improved abilities.

    Matching the fast, aggressive movements of an insect required a controller that could manage uncertainty and rapidly solve difficult optimization problems. Yet a system powerful enough to handle those calculations would ordinarily demand too much computing power for real-time flight, especially given the complex aerodynamics of such a lightweight machine.

    Software catches up with stronger hardware

    Chen’s group partnered with How’s team to overcome that conflict. Together, they created a two-step, AI-driven control system that combines the reliability needed for fast, complicated maneuvers with the efficiency required to operate in real time.

    “The hardware advances pushed the controller so there was more we could do on the software side, but at the same time, as the controller developed, there was more they could do with the hardware. As Kevin’s team demonstrates new capabilities, we demonstrate that we can utilize them,” How says.

    An expert planner trains the AI

    The first part of the system is a model-predictive controller. Using a mathematical model of the robot’s motion, it forecasts how the machine will behave and selects the best sequence of actions for safely following a planned route.

    This powerful planner can work out difficult maneuvers such as consecutive somersaults, sharp turns, and steep body tilts, although its calculations require substantial computing power. It also accounts for limits on the forces and torques the robot can generate, helping it maintain control and avoid collisions.

    Repeated flips are especially demanding because the robot must finish each rotation in the correct position and at the correct speed to begin the next one.

    “If small errors creep in, and you try to repeat that flip 10 times with those small errors, the robot will just crash. We need to have robust flight control,” How says.

    Instead of running the demanding planner directly during flight, the researchers used it as an expert teacher. Through imitation learning, it trained a “policy” built from a deep-learning model. This policy serves as the robot’s decision-making engine, determining where it should go and how it should move.

    Wind Tested Robot Performing Aerial Somersaults
    Even when wind disturbances threatened to push it off course, a speedy robot was agile enough to complete 10 consecutive somersaults in 11 seconds. Credit: Courtesy of the Soft and Micro Robotics Laboratory

    In effect, imitation learning transfers the capabilities of the computationally intensive planner into a faster AI model that can operate in real time.

    The challenge was generating enough useful training data to prepare the policy for aggressive maneuvers without producing more data than necessary.

    “The robust training method is the secret sauce of this technique,” How explains.

    During flight, the AI-driven policy receives the robot’s position and immediately produces control instructions, including the required thrust force and torques.

    Speed rises without sacrificing control

    Tests showed that the two-step system allowed the insect-scale robot to fly 447 percent faster and accelerate 255 percent more quickly than before. It completed 10 somersaults in 11 seconds while remaining within 4 or 5 centimeters of its intended path.

    “This work demonstrates that soft and microrobots, traditionally limited in speed, can now leverage advanced control algorithms to achieve agility approaching that of natural insects and larger robots, opening up new opportunities for multimodal locomotion,” says Hsiao.

    The robot also reproduced a movement known as a saccade. During this maneuver, insects tilt sharply, accelerate toward a location, and then pitch in the opposite direction to stop. This rapid change in speed helps them determine their position and maintain clear vision.

    “This bio-mimicking flight behavior could help us in the future when we start putting cameras and sensors on board the robot,” Chen says.

    Sensors could unlock independent flight

    A major next step will be equipping the microrobots with cameras and sensors so they can navigate outdoors without relying on an elaborate motion capture system.

    The researchers also plan to explore whether onboard sensors could help groups of robots coordinate their movements and avoid crashing into one another.

    “For the micro-robotics community, I hope this paper signals a paradigm shift by showing that we can develop a new control architecture that is high-performing and efficient at the same time,” says Chen.

    Reference: “Aerobatic maneuvers in insect-scale flapping-wing aerial robots via deep-learned robust tube model predictive control” by Yi-Hsuan Hsiao, Andrea Tagliabue, Owen Matteson, Suhan Kim, Tong Zhao, Jonathan P. How and YuFeng Chen, 3 December 2025, Science Advances.
    DOI: 10.1126/sciadv.aea8716

    This research is funded, in part, by the National Science Foundation (NSF), the Office of Naval Research, Air Force Office of Scientific Research, MathWorks, and the Zakhartchenko Fellowship.

    Never miss a breakthrough: Join the SciTechDaily newsletter.
    Follow us on Google and Google News.

    Artificial Intelligence Drones Machine Learning MIT Robotics
    Share. Facebook Twitter Pinterest LinkedIn Email Reddit

    Related Articles

    Solving the Tricky Challenges of Robotic Pizza-Making

    “Hey, Alexa! Are You Trustworthy?” – Human-Like Social Behaviors Improve Trust in Digital Assistants

    Computing for Ocean Environments: Bio-Inspired Underwater Devices & Swarming Algorithms for Robotic Vehicles

    New Machine-Learning System Gives Robots Social Skills

    New MIT Algorithm Helps Robots Collaborate to Get the Job Done

    MIT’s New Artificial Intelligence Algorithm Designs Soft Robots That Sense

    Showing Robots How to Do Your Chores – Automated Robots That Learn Just by Watching

    Innovative AI From MIT Helps Delivery Robots Find the Front Door [Video]

    Machine-Learning Models Capture Subtle Variations in Facial Expressions

    Leave A Reply Cancel Reply

    • Facebook
    • Twitter
    • Pinterest
    • YouTube

    Don't Miss a Discovery

    Subscribe for the Latest in Science & Tech!

    Trending News

    This Experimental Drug Repaired the Gut and Reversed Severe Fatty Liver Disease

    Physicists Just Turned a Black Hole Energy Theory Into Reality

    Stanford Scientists Solve a 252-Million-Year-Old Mass Extinction Mystery

    The Ancient Survival Mechanism Making Weight Loss So Difficult

    Just 3,000 Steps a Day May Help Protect the Brain From Alzheimer’s

    “Weird Clams” Reveal a New Invasion Along the U.S. Northeast Coast

    Archaeologists Decapitated Tutankhamun and Glued His Body Back Together

    New Black Hole Theory Solves a 50-Year-Old Problem

    Follow SciTechDaily
    • Facebook
    • Twitter
    • YouTube
    • Pinterest
    • Newsletter
    • RSS
    SciTech News
    • Biology News
    • Chemistry News
    • Earth News
    • Health News
    • Physics News
    • Science News
    • Space News
    • Technology News
    Recent Posts
    • New Study Reveals Herpesvirus Infection May Accelerate Alzheimer’s Disease
    • Microplastics Found in 84% of Serious Heart Attack Patients
    • The Oldest Known Right-Handed Animal May Have Lived 550 Million Years Ago
    • MIT’s Tiny Flying Robot Just Learned To Move Like a Real Insect
    • Why Are Bees Struggling To Survive This Summer?
    Copyright © 1998 - 2026 SciTechDaily. All Rights Reserved.
    • Science News
    • About
    • Contact
    • Editorial Board
    • Privacy Policy
    • Terms of Use

    Type above and press Enter to search. Press Esc to cancel.