Home Ecology Bird-inspired leg enables robots to jump into flight

Bird-inspired leg enables robots to jump into flight

Professor Monica Daley is a co-author and biology contributor for this Nature paper on RAVEN: a bird-inspired multi-modal robot that uses its legs to propel itself for take-off for flight.

  • 04 December 2024

  • A bird’s ability to transition swiftly from walking to flying has been emulated in the design of a simple robotic leg. The feat is a win for drone design, and one that could reveal why some birds’ legs make up so large a fraction of their weight.
  • Aimy A. Wissa

Many organisms can transition smoothly from one mode of locomotion to another, often by using the same body parts. For example, flying fish use their fins to steer underwater, and these same structures turn into wings when the fish begin to glide above the water’s surface1. Such versatility is an endless source of inspiration for engineers, but using the same locomotor modules to perform different movements can create trade-offs, because these manoeuvres often have opposing requirements2.Read the paper: Fast ground-to-air transition with avian-inspired multifunctional legs

Writing in Nature, Shin et al.3 report the design of a bird-inspired robotic leg that has many locomotion modes — and can transition between them with remarkable speed and efficiency.

Little is known about the evolutionary pressures and mechanistic trade-offs that govern transitions between different modes of locomotion from a biological perspective. Furthermore, some organisms that show multimodal locomotion are among the hardest to study, especially when they are transitioning from one mode to another — from walking to burrowing, for example, or from flying to swimming.

This gap in understanding has hampered efforts to emulate multimodal locomotion in engineered systems — a feature that would benefit robots in several settings, including when moving over rough terrain or navigating obstacles much larger than the machines themselves. Although some researchers have succeeded in building robotic systems that are capable of several modes of locomotion47, most are limited to just two. And there is no generalized framework that would enable engineers to transfer these abilities to other systems. For these reasons, multimodal locomotion presents exciting scientific opportunities in both engineering and biology.Embrace wobble to level flight without a horizon

Shin et al. took inspiration from the anatomy of birds to design a multifunctional robotic leg that enables RAVEN, the authors’ uncrewed aerial vehicle (UAV), to walk, hop, jump and fly. Adding legs to such vehicles expands their range of abilities substantially, allowing them to navigate complex environments. It also enables them to reduce their take-off time and distance requirements, because they can jump into flight instead of needing to be launched or dropped. However, the design of such legs can add complexity and increase energy costs — a problem that the team circumvented by coming up with a simplified design.

Instead of building an exact replica of a bird leg, featuring life-like joints and modes of motion, Shin et al. devised a structure containing two leg segments that are linked to a foot through a toe joint (Fig. 1a). The two leg segments are connected by a joint that simulates an ankle, and attached to the UAV through a hip-like structure. By reducing the complexity of a real bird leg in this way, the authors were able to keep the structure light and easy to control. And by positioning the control modules in the hip joint, they could concentrate most of the mass in the upper leg, thus minimizing the energy expended.

figure 1
Figure 1 | A robotic leg capable of multimodal locomotion. Shin et al.3 took inspiration from birds’ legs to design a robotic limb that can switch between modes of locomotion. a, Unlike a bird leg, the authors’ structure comprises just two segments connected by joints to simulate an ankle, hip and foot. b, An uncrewed aerial vehicle equipped with this leg can take flight simply by jumping from a standing position. The launch mechanism mimics birds’ ability to store elastic energy by flexing their legs, and then to rapidly release this energy by extending their legs. (Adapted from Extended Data Fig. 1 in ref. 3.)

The mechanism through which the UAV launches was also devised through analogy with birds (Fig. 1b). Avian take-off involves elastic energy being stored (through flexing the leg) and then released rapidly (by extending the leg). Shin et al. mimicked this release mechanism by adding a spring to the ankle joint of their bird-inspired leg. The foot is also equipped with a spring, which helps the leg to change direction, and to endure destabilizing movements while walking or launching into flight mode.

Using this simplified design, the authors were able to make their UAV exhibit complex terrestrial locomotion such as hopping, walking and jumping onto obstacles. They also investigated various take-off strategies, including three that used a front propeller. In the first approach, the craft was launched from a standing position; in the second, by falling from a height; and in the third, by jumping. Intriguingly, the team concluded that transitioning into flight using only a jump was almost as fast as jumping with assistance from the propeller — and it was faster than taking off from standing or by falling. The jump-to-flight strategy (without the propeller) was also more energy-efficient than the other three methods.The 50th anniversary of a key paper on how bird flight evolved

Shin and colleagues’ work is more than simply a viable solution to achieving multimodal locomotion for UAVs. Indeed, RAVEN can be used to provide insights about how birds transition between being on the ground and being in the air. The authors’ experiments on RAVEN helped them to form a hypothesis about why some birds that fly have legs that take up nearly 30% of their overall body mass. The results suggest that these legs might be energetically ‘worth it’, because they enable fast and energy-efficient take-off. This ability is particularly crucial for birds that frequently perform ground-to-air transitions, such as birds of prey.

Perhaps even more importantly, the findings confirm that engineering and biology form an interdisciplinary two-way street8,9. In one direction, known as bioinspired design, natural solutions can inform strategies for building mechanical systems. In the other direction, referred to as engineering-enabled biology, tools such as robotic model organisms are used to address fundamental biological questions that are difficult to answer through direct observation of a natural system.

Multimodal locomotion is an area that can benefit immensely from this transdisciplinary approach — and Shin and colleagues’ clever robotic design is testament to this potential. As research continues to grow in this area, I anticipate the emergence of bioinspired multimodal robotic systems that can be deployed in new and complex environments. Such systems will also aid in biological discoveries by uncovering the physical principles that organisms use to transition seamlessly between modes of locomotion.

Nature 636, 48-49 (2024)

doi: https://doi.org/10.1038/d41586-024-03845-w

References

  1. Davenport, J. Rev. Fish Biol. Fish. 4, 184–214, (1994).Article Google Scholar 
  2. Siddall, R. & Kovač, M. Bioinspir. Biomim. 9, 031001 (2014).Article PubMed Google Scholar 
  3. Shin, W. D., Phan, H.-V., Daley, M. A., Ijspeert, A. J. & Floreano, D. Nature 636, 86–91 (2024).Article Google Scholar 
  4. Chen, Y. et al. Sci. Robot. 2, eaao5619 (2017).Article PubMed Google Scholar 
  5. Desbiens, A. L., Pope, M. T., Christensen, D. L., Hawkes, E. W. & Cutkosky, M. R. Bioinspir. Biomim. 9, 025009 (2014).Article PubMed Google Scholar 
  6. Ramirez, J. P. & Hamaza, S. Adv. Intell. Syst. https://doi.org/10.1002/aisy.202300327 (2023).Article Google Scholar 
  7. Baines, R. et al. Nature 610, 283–289 (2022).Article PubMed Google Scholar 
  8. Wissa, A., Alleyne, M., Barley, W. C. & Suarez, A. V. Integr. Comp. Biol. 62, 1147–1152 (2022).Article Google Scholar 
  9. Ijspeert, A. J. Science 346, 196–203 (2014).Article PubMed Google Scholar 

Download references

Share This Post

Post Navigation