Untethered flight of an insect-sized flapping-wing microscale aerial vehicle


Heavier-than-air flight at any scale is energetically expensive. This is greatly exacerbated at small scales and has so far presented an insurmountable obstacle for untethered flight in insect-sized (mass less than 500 milligrams and wingspan less than 5 centimetres) robots. These vehicles1,2,3,4 thus need to fly tethered to an offboard power supply and signal generator owing to the challenges associated with integrating onboard electronics within a limited payload capacity. Here we address these challenges to demonstrate sustained untethered flight of an insect-sized flapping-wing microscale aerial vehicle. The 90-milligram vehicle uses four wings driven by two alumina-reinforced piezoelectric actuators to increase aerodynamic efficiency (by up to 29 per cent relative to similar two-wing vehicles5) and achieve a peak lift-to-weight ratio of 4.1 to 1, demonstrating greater thrust per muscle mass than typical biological counterparts6. The integrated system of the vehicle together with the electronics required for untethered flight (a photovoltaic array and a signal generator) weighs 259 milligrams, with an additional payload capacity allowing for additional onboard devices. Consuming only 110–120 milliwatts of power, the system matches the thrust efficiency of similarly sized insects such as bees7. This insect-scale aerial vehicle is the lightest thus far to achieve sustained untethered flight (as opposed to impulsive jumping8 or liftoff9).

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Fig. 1: The integrated vehicle and flight performance.
Fig. 2: Vehicle design.
Fig. 3: Drive electronics.
Fig. 4: Measured drive signals.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.


  1. 1.

    Ma, K. Y., Felton, S. M. & Wood, R. J. In IEEE Int. Conf. on Intelligent Robots and Systems (IROS) 1133–1140 (IEEE/RSJ, 2012).

  2. 2.

    Zou, Y., Zhang, W. & Zhang, Z. Liftoff of an electromagnetically driven insect-inspired flapping-wing robot. IEEE Trans. Robot. 32, 1285–1289 (2016).

    Google Scholar 

  3. 3.

    Balasubramanian, S., Chukewad, Y. M., James, J. M., Barrows, G. L. & Fuller, S. B. An insect-sized robot that uses a custom-built onboard camera and a neural network to classify and respond to visual input. In 7th IEEE Int. Conf. on Biomedical Robotics and Biomechatronics (Biorob) 1297–1302 (IEEE, 2018).

  4. 4.

    Drew, D. S. & Pister, K. S. In IEEE Int. Conf. on Manipulation, Automation and Robotics at Small Scales (MARSS) 1–5 (IEEE, 2017).

  5. 5.

    Jafferis, N. T., Graule, M. A. & Wood, R. J. In 2016 IEEE Int. Conf. on Robotics and Automation (ICRA) 3234–3241 (IEEE, 2016).

  6. 6.

    Marden, J. H. Maximum lift production during takeoff in flying animals. J. Exp. Biol. 130, 235–258 (1987).

    Google Scholar 

  7. 7.

    Nachtigall, W., Hanauer-Thieser, U. & Mörz, M. Flight of the honey bee. VII: Metabolic power versus flight speed relation. J. Comp. Physiol. B 165, 484–489 (1995).

    Google Scholar 

  8. 8.

    Churaman, W. A., Currano, L. J., Morris, C. J., Rajkowski, J. E. & Bergbreiter, S. The first launch of an autonomous thrust-driven microrobot using nanoporous energetic silicon. J. Microelectromech. Syst. 21, 198–205 (2012).

    Google Scholar 

  9. 9.

    James, J., Iyer, V., Chukewad, Y., Gollakota, S. & Fuller, S. B. In IEEE Int. Conf. on Robotics and Automation (ICRA) 1–8 (IEEE, 2018).

  10. 10.

    Keennon, M., Klingebiel, K. & Won, H. In 50th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition 588 (AIAA, 2012).

  11. 11.

    Rosen, M. H., le Pivain, G., Sahai, R., Jafferis, N. T. & Wood, R. J. In IEEE Int. Conf. on Robotics and Automation (ICRA). 3227–3233 (IEEE, 2016).

  12. 12.

    de Croon, G., de Clercq, K., Ruijsink, R., Remes, B. & de Wagter, C. Design, aerodynamics, and vision-based control of the delfly. Int. J. Micro Air Veh. 1, 71–97 (2009).

    Google Scholar 

  13. 13.

    Karásek, M., Muijres, F. T., de Wagter, C., Remes, B. D. & de Croon, G. C. A tailless aerial robotic flapper reveals that flies use torque coupling in rapid banked turns. Science 361, 1089–1094 (2018).

    Google Scholar 

  14. 14.

    Hines, L. Design and Control of a Flapping Flight Micro Aerial Vehicle. PhD thesis, Carnegie Mellon Univ. (2014); https://www.ri.cmu.edu/publications/design-and-control-of-a-flapping-flight-micro-aerial-vehicle/.

  15. 15.

    Steltz, E., Avadhanula, S. & Fearing, R. S. In IEEE Int. Conf. on Intelligent Robots and Systems (IROS) 3987–3992 (IEEE/RSJ, 2007).

  16. 16.

    Anderson, M. L. Design and Control of Flapping Wing Micro Air Vehicles. PhD thesis, Air Force Institute of Technology Wright-Patterson AFB OH School of Engineering and Management (2011); https://apps.dtic.mil/docs/citations/ADA549053.

  17. 17.

    Smith, G. L. et al. PZT-based piezoelectric MEMS technology. J. Am. Ceram. Soc. 95, 1777–1792 (2012).

    Google Scholar 

  18. 18.

    Jafferis, N. T., Smith, M. J. & Wood, R. J. Design and manufacturing rules for maximizing the performance of polycrystalline piezoelectric bending actuators. Smart Mater. Struct. 24, 065023 (2015).

    Google Scholar 

  19. 19.

    Karpelson, M., Wei, G.-Y. & Wood, R. J. Driving high voltage piezoelectric actuators in microrobotic applications. Sens. Actuators A 176, 78–89 (2012).

    Google Scholar 

  20. 20.

    Steltz, E., Seeman, M., Avadhanula, S. & Fearing, R. S. In Int. Conf. on Intelligent Robots and Systems 1322–1328 (IEEE/RSJ, 2006).

  21. 21.

    Goldberg, B. et al. Power and control autonomy for high-speed locomotion with an insect-scale legged robot. IEEE Robot. Autom. Lett. 3, 987–993 (2018).

    Google Scholar 

  22. 22.

    Duduta, M., de Rivaz, S., Clarke, D. R. & Wood, R. J. Ultralightweight, high power density lithium-ion batteries. Batteries Supercaps 1, 131–134 (2018).

    Google Scholar 

  23. 23.

    Brühwiler, R. et al. In Int. Conf. on Intelligent Robots and Systems 5727–5733 (IEEE/RSJ, 2015).

  24. 24.

    Hollar, S., Flynn, A., Bellew, C. & Pister, K. In 16th Ann. Int. Conf. on Micro Electro Mechanical Systems (MEMS-03) 706–711 (IEEE, 2003).

  25. 25.

    Boucher, R. J. Sunrise, the world’s first solar-powered airplane. J. Aircr. 22, 840–846 (1985).

    Google Scholar 

  26. 26.

    Fuller, S. B. Four wings: an insect-sized aerial robot with steering ability and payload capacity for autonomy. IEEE Robot. Autom. Lett. 4, 570–577 (2019).

    Google Scholar 

  27. 27.

    van Breugel, F., Regan, W. & Lipson, H. From insects to machines. IEEE Robot. Autom. Mag. 15, 68–74 (2008).

    Google Scholar 

  28. 28.

    Teoh, Z. E. et al. In Int. Conf. on Intelligent Robots and Systems 3209–3216 (IEEE/RSJ, 2012).

  29. 29.

    Graule, M., et al. Perching and takeoff of a robotic insect on overhangs using switchable electrostatic adhesion. Science 352, 978–982 (2016).

    Google Scholar 

  30. 30.

    Fuller, S. B., Helbling, E. F., Chirarattananon, P. & Wood, R. J. Using a MEMS gyroscope to stabilize the attitude of a fly-sized hovering robot. In International Micro Air Vehicle (IMAV) Conference and Competition (Delft University of Technology (TU Delft), 2014).

  31. 31.

    Helbling, E. F., Fuller, S. B. & Wood, R. J. In Robotics Research 57–69 (Springer, 2018).

  32. 32.

    Zhang, X. et al. In IEEE Symp. on VLSI Circuits C152–C153 (IEEE, 2015).

  33. 33.

    Jafferis, N. T., Lok, M., Winey, N., Wei, G.-Y. & Wood, R. J. Multilayer laminated piezoelectric bending actuators: design and manufacturing for optimum power density and efficiency. Smart Mater. Struct. 25, 055033 (2016).

    Google Scholar 

  34. 34.

    Fuller, S. B. et al. Stabilizing air dampers for hovering aerial robotics: design, insect-scale flight tests, and scaling. Auton. Robots 41, 1555–1573 (2017).

    Google Scholar 

  35. 35.

    Jayaram, K., Jafferis, N. T., Doshi, N., Goldberg, B. & Wood, R. J. Concomitant sensing and actuation for piezoelectric microrobots. Smart Mater. Struct. 27, 065028 (2018).

    Google Scholar 

Download references


We thank R. Peña Velasco for assistance in PCB fabrication. This work is supported by the National Science Foundation (award numbers 1514306 and 1724197), the Office of Naval Research (award number N000141712614), and the Wyss Institute for Biologically Inspired Engineering. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

Reviewer information

Nature thanks Kenny Breuer, Kristofer Pister and Franck Ruffier for their contribution to the peer review of this work.

Author information




N.T.J. designed, fabricated and characterized the vehicle; assembled the photovoltaic array; conceived the drive waveform; and performed the system analysis. E.F.H. and M.K. designed and fabricated the power electronics and implemented the control architecture for the drive waveform. E.F.H. and N.T.J. assembled the integrated vehicle, characterized system performance and executed the flight tests. E.F.H. and N.T.J. wrote the paper. All authors provided feedback.

Corresponding author

Correspondence to Noah T. Jafferis.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 Actuator cross-sectional diagrams.

a, The actuators developed in ref. 18. b, The alumina-reinforced actuators used in this work. The blue arrows indicate the bending direction of the actuators.

Extended Data Fig. 2 Lift force measurement via vertical acceleration.

The vehicle’s altitude versus time when operated at 210 V, 165 Hz, and carrying a payload of 245 mg (for a total mass of 335 mg). The voltage is increased from zero to 210 V from t = −0.1 s to t = 0 s, and remains at 210 V for the remainder of the trial. The curvature of this trajectory shows that the lift is 370 mg for this operating condition. The lift is calculated from a least-squares fit to the circled data points, which occur while the vehicle’s deviation from vertical (labelled ‘angle’ in the plot) is small enough to cause errors <1%; correcting for the angle error gives the same lift value over the entire flight time.

Extended Data Fig. 3 Illustration of torque implementation.

ac, Pitch torque (a), roll torque (b) and yaw torque (c) (all depicted with ‘top-down’ views of the wing strokes), in the X-Wing versus the previous two-wing design. The red and green squares in the X-Wing design indicate that each pair of wings is fixed at 90 degrees apart. Although the four-wing design brings the mean centre of pressure of each wing pair closer to the body for pitch and roll (indicated by the length of the dashed lines), it also produces greater forces for the same flapping motion, allowing for similar control authority.

Extended Data Fig. 4 Power supply.

a, Six-cell photovoltaic array. b, Plot of the power obtained from this solar array at varying distances from our light source (the measurement precision is approximately ±5 mW). The power at 1 Sun is approximately 46 mW.

Extended Data Fig. 5

Drive waveform switching control diagram.

Extended Data Fig. 6 Trajectories for the untethered flights.

Z is altitude, X and Y are lateral displacements, and ‘Angle’ is the projection of the vehicle’s orientation onto the XZ plane (that is, the camera view), with vertical being zero. The light turns on at t = 0, and the vertical dashed line indicates when it has reached full intensity. a, The flight shown in Supplementary Video 9 (the lowest-power flight). b, The flight shown in Supplementary Video 4. Depth (Y) is estimated by measuring the fractional change in the peak width of the flexboard as our integrated vehicle spins about the yaw axis (that is, twice every yaw period, we measure the width of the flexboard), along with the starting distance from the vehicle to the camera (approximately 90 cm). The error bars on Y are due to the pixel resolution error (±0.2 mm) amplified by the ratio of the camera distance to the flexboard width. The lift force is estimated from the vertical acceleration (z″) of the vehicle during the first 3 cm of vertical flight (indicated by the circled data points), before the vehicle has noticeably tilted or moved laterally (note that for the flight in b, our lift estimate is based on data before the light has reached full intensity, because the vehicle has already moved and tilted noticeably by then).

Extended Data Fig. 7 System trade-off analysis.

This series of plots compares predictions for the required power (Pn) and the power available at 1 Sun (Pa) (power density 0.76 W g−1) (top row); the mean light intensity required to fly (in number of Suns) (middle row); and the area of solar cells that can be carried (As) divided by the area ‘swept’ by the wings (π(0.5S)2) (bottom row). These quantities are plotted as functions of overall vehicle scaling for vehicles with two or four wings and two different transmission ratios. Plots are shown for three actuator widths: 1× = 1.125 mm (black), 2× = 2.250 mm (red) and 3× = 3.375 mm (blue). The circled and squared dots correspond to predictions for SDAB15 and the vehicle described in this work, respectively. The steps in the plots are due to the discrete size of available solar cells.

Extended Data Table 1 Vehicle characterization
Extended Data Table 2 Components of the signal-generation electronics and their corresponding masses

Supplementary information

Vehicle flapping kinematics

Video 1 . The operating condition is 210 V and 165 Hz. The wing stroke is 88º peak to peak at this condition (see Extended Data Table 1b for more details).

Vehicle weight-lifting test

Video 2 . The vehicle is flown open-loop at the same operating condition as in Supplementary Video 1. An additional mass of 245 mg is affixed to the vehicle, resulting in a total mass of 335 mg. Based on the vehicle’s vertical acceleration, its lift is calculated to be 370 mg.

Flapping kinematics comparison

Video 3 . In the left-hand video, an external high-voltage amplifier directly provides the drive signals (193 V and 173 Hz) supplied to the vehicle. In the right-hand video, the drive signals are produced by the on-board drive electronics (with an external 4.8 V power supply in place of the photovoltaic array). No difference is seen in the kinematics.

Untethered flight

Video 4 . The integrated vehicle successfully takes off and performs a sustained untethered flight. The light source provides a light intensity of approximately three Suns directly above the vehicle, outside the field of view. The Kevlar thread above the solar array is for vehicle safety (it provides neither current nor support during flight).

Vehicle flapping kinematics

Video 5 . The operating condition is 200 V and 170 Hz. The wing stroke is 77.5º peak to peak at this condition (see Extended Data Table 1b for more details).

Vehicle flapping kinematics

Video 6 . The operating condition is 190 V and 173 Hz. The wing stroke is 68.8º peak to peak at this condition (see Extended Data Table 1b for more details).

Vehicle weight-lifting test

Video 7 . The vehicle is flown open-loop at the same operating condition as in Supplementary Video 5. An additional mass of 224 mg is affixed to the vehicle, resulting in a total mass of 314 mg. Based on the vehicle’s vertical acceleration, its lift is calculated to be 340 mg.

Vehicle weight-lifting test

Video 8 . The vehicle is flown open-loop at the same operating condition as in Supplementary Video 6. An additional mass of 224 mg is affixed to the vehicle, resulting in a total mass of 314 mg. Based on the vehicle’s vertical acceleration, its lift is calculated to be 325 mg.

Untethered flight

Video 9 . This untethered flight was performed with the minimum initial available peak power (150 mW) for a successful flight.

Pitch and roll demonstration

Video 10 . The vehicle is mounted in one of two orientations on a magnetic pin such that it is free to rotate about either the pitch (top row) or roll (bottom row) axes. Open-loop torque commands are applied to implement (from left to right) negative, zero, and positive pitch and roll.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Jafferis, N.T., Helbling, E.F., Karpelson, M. et al. Untethered flight of an insect-sized flapping-wing microscale aerial vehicle. Nature 570, 491–495 (2019). https://doi.org/10.1038/s41586-019-1322-0

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing