A Caltech Library Service

Neural-Fly enables rapid learning for agile flight in strong winds

O’Connell, Michael and Shi, Guanya and Shi, Xichen and Azizzadenesheli, Kamyar and Anandkumar, Anima and Yue, Yisong and Chung, Soon-Jo (2022) Neural-Fly enables rapid learning for agile flight in strong winds. Science Robotics, 7 (66). Art. No. eabm6597. ISSN 2470-9476. doi:10.1126/scirobotics.abm6597.

[img] PDF (Sections S1 to S8; Figs. S1 to S4; Tables S1 to S3; References (52–57)) - Supplemental Material
See Usage Policy.


Use this Persistent URL to link to this item:


Executing safe and precise flight maneuvers in dynamic high-speed winds is important for the ongoing commoditization of uninhabited aerial vehicles (UAVs). However, because the relationship between various wind conditions and its effect on aircraft maneuverability is not well understood, it is challenging to design effective robot controllers using traditional control design methods. We present Neural-Fly, a learning-based approach that allows rapid online adaptation by incorporating pretrained representations through deep learning. Neural-Fly builds on two key observations that aerodynamics in different wind conditions share a common representation and that the wind-specific part lies in a low-dimensional space. To that end, Neural-Fly uses a proposed learning algorithm, domain adversarially invariant meta-learning (DAIML), to learn the shared representation, only using 12 minutes of flight data. With the learned representation as a basis, Neural-Fly then uses a composite adaptation law to update a set of linear coefficients for mixing the basis elements. When evaluated under challenging wind conditions generated with the Caltech Real Weather Wind Tunnel, with wind speeds up to 43.6 kilometers/hour (12.1 meters/second), Neural-Fly achieves precise flight control with substantially smaller tracking error than stateof-the-art nonlinear and adaptive controllers. In addition to strong empirical performance, the exponential stability of Neural-Fly results in robustness guarantees. Last, our control design extrapolates to unseen wind conditions, is shown to be effective for outdoor flights with only onboard sensors, and can transfer across drones with minimal performance degradation.

Item Type:Article
Related URLs:
URLURL TypeDescription ItemCode
O’Connell, Michael0000-0001-6681-8823
Shi, Guanya0000-0002-9075-3705
Shi, Xichen0000-0002-5366-9256
Azizzadenesheli, Kamyar0000-0001-8507-1868
Anandkumar, Anima0000-0002-6974-6797
Yue, Yisong0000-0001-9127-1989
Chung, Soon-Jo0000-0002-6657-3907
Additional Information:© 2022 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. Submitted 11 October 2021; Accepted 12 April 2022; Published 4 May 2022. A.A. is also affiliated with NVIDIA Corporation, and Y.Y. is also with associated Argo AI. K.A. is currently affiliated with Purdue University. We thank J. Burdick and J.-J. E. Slotine for their helpful discussions. We thank M. Anderson for help with configuring the quadrotor platform, and M. Anderson and P. Spieler for help with hardware troubleshooting. We also thank N. Badillo and L. Pabon Madrid for help in experiments. This research was developed with funding from the Defense Advanced Research Projects Agency (DARPA). This research was also conducted in part with funding from Raytheon Technologies. The views, opinions, and/or findings expressed are those of the authors and should not be interpreted as representing the official views or policies of the Department of Defense or the U.S. Government. The experiments reported in this article were conducted at Caltech’s Center for Autonomous Systems and Technologies (CAST). Author contributions: S.-J.C. and Y.Y. directed the research activities. G.S. and M.O. designed and implemented the metalearning algorithm under the guidance of Y.Y., K.A., A.A., and S.-J.C., while the last-layer adaptation idea was started with a discussion by G.S., M.O., X.S., and S.-J.C. M.O. and G.S. designed and implemented the adaptive control algorithm with inputs from S.-J.C. and X.S. M.O. and G.S. performed experiments and evaluated the results. M.O. conducted the theoretical analysis of the meta-learning based adaptive controller with input from S.-J.C., G.S., and X.S. G.S. analyzed the learning algorithm with feedback from Y.Y., K.A., A.A., and S.-J.C. G.S. and M.O. created all the figures and videos with input from the other authors. All authors prepared the manuscript. The authors declare that they have no competing interests. Data and materials availability: All data needed to evaluate the conclusions in the article are present in the article or in the Supplementary Materials. We have provided the machine learning model training code, training data, and experimental data at
Group:Center for Autonomous Systems and Technologies (CAST), GALCIT
Funding AgencyGrant Number
Defense Advanced Research Projects Agency (DARPA)UNSPECIFIED
Raytheon CompanyUNSPECIFIED
Issue or Number:66
Record Number:CaltechAUTHORS:20220505-792409800
Persistent URL:
Official Citation:Neural-Fly enables rapid learning for agile flight in strong winds. Michael O’Connell, Guanya Shi, Xichen Shi, Kamyar Azizzadenesheli, Anima Anandkumar, Yisong Yue, Soon-Jo Chung. Sci. Robot., 7 (66), eabm6597; DOI: 10.1126/scirobotics.abm6597
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:114603
Deposited By: Tony Diaz
Deposited On:05 May 2022 23:33
Last Modified:05 May 2022 23:33

Repository Staff Only: item control page