Published: 19 Nov 2025 | Last Updated: 19 Nov 2025 16:01:18

A new study led by the Royal Veterinary College (RVC) and Imperial College London has revealed how dragonflies use a small number of strategically positioned sensors on their wings to monitor wing shape in real time. This simple but highly effective biological system enables the insects to remain exceptionally stable or manoeuvre in flight, despite having comparatively tiny brains. The findings of this research offer important new insights for bioengineering that could inspire future aircraft, robotics and autonomous vehicle design.

As is the case with most insects, dragonfly wings can bend and twist in many ways. Yet, little research has been undertaken to understand how these insects monitor these changes quickly enough to stay in control. The wings contain no muscles but gather information from sensors on the wing veins about how they deform during flight.

The lead author, Dr Alexandra Yarger, Research Associate at Imperial College London in the lab of Dr Huai-Ti Lin, Associate Professor of Neuromechanics and Bioinspired Technologies, recorded neural activity as the wings fluttered and were stimulated by air flows to mimic natural bending and twisting. Dr Yarger combined these measurement with detailed mechanical simulations of wing behaviour based on high-speed, multi-camera filming of free-flying insects led by Professor Richard Bomphrey, Professor of Comparative Biomechanics at the RVC and Dr Simon Walker, Associate Professor in Comparative Biomechanics at the University of Leeds. This integrative approach revealed how the timing of sensory activity encodes aerodynamic and inertial forces and how additional sensors are recruited when the wing is perturbed.

The results show that wing structure plays a key role in constraining how the wing can deform, creating a close alignment between mechanical design and sensor placement. While the wide range of possible wing configurations might have presented a significant challenge for the nervous system to process, the findings of this study demonstrate that, in practice, almost all wing deformation during normal flight falls into a small number of dominant patterns. In fact, the researchers found that approximately 99% of wing displacement can be explained by three principal components – bend, twist and camber.

Importantly, their analysis found that wing strain sensors are positioned exactly where changes in strain patterns are best detected, meaning only a few signals are needed to represent the overall wing state comprehensively, simplifying the task for the brain. This provides a compelling example of morphological computation, in which the physical structure of an organism naturally simplifies the sensory and computational load on the nervous systems. Under normal, unperturbed flight conditions, only a few sensors are required to encode the full wing state. However, when the wing is perturbed, additional sensors are recruited to capture the more complex deformations.

These findings help explain how insects with relatively simple neural architectures can monitor and react to complex aerodynamic conditions with remarkable speed. However, beyond biological insight, the study highlights principles that could be replicated in engineering, including applications for flexible aircraft wings, flapping-wing drones and other lightweight, deformable technologies. By designing structures that limit unnecessary deformation and positioning sensors at the most informative locations, engineered systems could achieve reliable control with fewer sensors and reduced computational demand.

Dr Alexandra Yarger, Research Associate at Imperial College London, said:

“The integration of wing structure and sensor placement is an elegant example of morphological computation - where the body itself filters and organises sensory information before it even reaches the nervous system.

“We found that the seemingly complex deformations can be reduced to a few dominant features (mainly bending and twisting) that allow dragonflies to effectively and efficiently monitor their wings.

“With just a few well-placed sensors, dragonflies can effectively monitor their wings in real time, providing a robust and efficient strategy for flight control.

“Dragonflies have had hundreds of millions of years to solve many of the same problems engineers still face, so rather than inventing entirely new solutions, we’re learning from a system shaped by evolution to be successful.”

Richard Bomphrey, Professor of Comparative Biomechanics at the RVC, said:

“Using a wide range of techniques – ranging from high-speed, multi-camera filming to capture how dragonfly wings change shape during flight, to recording the activity of neurons that fire in response to certain combinations of bending or twisting – we have observed an elegant evolved system that helps insects to fly so effectively despite comparatively tiny brains.”

Huai-Ti Lin, Associate Professor of Neuromechanics and Bioinspired Technologies at Imperial College London, said:

“It’s been ten years since I first placed an electrode into a dragonfly wing and started imagining the sensory world hidden inside it. The journey since then has been incredibly inspiring, thanks to the hard work and creativity of a fantastic multidisciplinary team. Our findings on how a flexible wing encodes sensory information really show how mechanics shape sensorimotor control. It’s now my go-to example whenever I need to explain what neuro-mechanics is.”

This study was supported by the Biotechnology and Biological Sciences Research Council, the Ansys Academic Research Partner Programme, the Royal Society Research Fellows Enhancement Award and The Grass Foundation.

Notes to Editors

References

A.M. Yarger, M. Maeda, I. Siwanowicz, H. Kajiyama, S.M. Walker, R.J. Bomphrey, & H. Lin, Structural dynamics and neural representation of wing deformation, Proc. Natl. Acad. Sci. U.S.A. 122 (46) e2518032122, https://doi.org/10.1073/pnas.2518032122 (2025).

The full paper can be accessed at: https://doi.org/10.1073/pnas.2518032122

For media enquiries, please contact:

• rvc@plmr.co.uk
• Press Line: 0800 368 9520

About the RVC

• The Royal Veterinary College (RVC) is the UK's largest and longest established independent veterinary school and is a Member Institution of the University of London.
• It is one of the few veterinary schools in the world that hold accreditations from the RCVS in the UK (with associated recognition from the AVBC for Australasia, the VCI for Ireland and the SAVC for South Africa), the EAEVE in the EU, and AVMA in the USA and Canada.
• The RVC is ranked as the top veterinary school in the world in the QS World University Rankings by subject, 2025.
• The RVC offers undergraduate and postgraduate programmes in veterinary medicine, veterinary nursing and biological sciences.
• The RVC is a research-led institution, with 88% of its research rated as internationally excellent or world class in the Research Excellence Framework 2021.
• The RVC provides animal owners and the veterinary profession with access to expert veterinary care and advice through its teaching hospitals and first opinion practices in London and Hertfordshire.