Inspired by Manta Rays, “Butterfly Bot” is the Fastest Swimming Soft Bot Yet

November 30, 2022 by Aaron Carman

Taking the plunge into swimming robotics, North Carolina State researchers have developed a device inspired by both the butterfly stroke and manta rays.

In an era of increased robotic proliferation, North Carolina State researchers have developed a “Butterfly Bot” to allow next-generation robots to operate in water. Despite its name, the Butterfly Bot is not an airborne automaton but a highly efficient swimming robot modeled after the swim patterns of manta rays. 


A system-level diagram of the Butterfly Bot

A system-level diagram of the Butterfly Bot highlights the main components of the robot. The soft body contains an actuator that can efficiently propel the robot thanks to the wings and fins. Image used courtesy of NC State


The Butterfly Bot, named after the butterfly stroke, boasts a top speed four times faster than previous swimming robots, all the while using a relatively simple design. The NC State group developed two prototypes, each with its own strengths. This article gives a deeper look at the technology behind the Butterfly Bot.


Float Like a Butterfly

Taking inspiration from the manta ray, a small yet speedy sea creature, the NC State research group led by Dr. Jie Yin noticed that the efficient “flapping” motion seen so often in nature is not nearly as efficient in bio-inspired designs. While marine animals can often achieve speeds greater than one body length per second (BL/s), previous designs could not surpass this milestone. In addition, they were less efficient, making them impractical to use.

With Occam’s razor in mind, the NC State group developed an elegant swimming robot that utilizes the “flapping” motion of manta rays. Instead of relying on distributed complex electrical or hydraulic systems, the Butterfly Bot relies on bistable flexible wings to produce forward motion. Bistable wings, much like a hairclip, have two stable states. By adding a sufficient external stimulus, it’s possible to “switch” the states, which can often accompany a rapid release of the stored energy.


The relationship between the soft body bending angle and wing rotation angle

The relationship between the soft body bending angle and wing rotation angle demonstrates the bistable principle used in the Butterfly Bot. Once a critical angle is reached, the wings snap into a new stable state. Image used courtesy of NC State


Bistable wings offer several advantages to the Butterfly Bot, both in system complexity and raw performance. Regarding system complexity, the wings may be assembled quite easily, requiring only one glued joint per wing. Driving the wings is trivial, requiring a central drive unit that can control the position of multiple wings simultaneously (although not independently). In the first Butterfly Bot, the wings experienced a peak velocity of 6.6 m/s and a peak acceleration of 1.49 x 103 m/s2.


Swim Like a Manta Ray

To further mimic the motion of a manta ray, a flexible fin was added to the backside of the wing to allow the “flap” to continue down the length of the fin, producing forward motion and allowing the Butterfly Bot to not only swim but do so efficiently.

To assess the effectiveness of the Butterfly Bot, two main metrics were used: the top speed and the Strouhal number. Top speed is typically measured in BL/s since a larger body will naturally exhibit a larger peak velocity, while the Strouhal number is a dimensionless measure that represents power efficiency. Strouhal numbers in nature are typically between 0.2 and 0.4, making this range the target for the study. 


The frequency-speed/efficiency relationship

The frequency-speed/efficiency relationship demonstrates that maximum efficiency is achieved at ~0.67 Hz, but an increased speed may be observed by increasing actuation frequency. Image used courtesy of NC State


For the first Butterfly Bot, the researchers observed a top speed of 3.74 BL/s with an actuation pressure of 55 kPa and 1 Hz actuation frequency. It is worth noting that different actuation frequencies will naturally produce different speeds and energy efficiencies (much how like we achieve a higher speed when sprinting but exhaust energy much faster). The peak efficiency occurred at an actuation frequency of 0.67 Hz, with an associated speed of 3.4 BL/s.

Up to this point, researchers have only used a single actuator to control both wings. This creates the simplest system architecture while constraining the motion to a line. However, the NC State group wanted to enable the Butterfly Bot to turn. As such, the researchers assembled a second bot, this time with two actuators, to independently control the wings. By flapping a single wing, the robot can turn, while flapping simultaneously allows the bot to move forward with enhanced maneuverability.


The Sea: The Final Frontier for Robotics

In true scientific fashion, the NC State group realized that, despite the impressive achievements of the Butterfly Bot, there is still more to do to improve their aquatic creation. One of the main limiting factors to speed is the bandwidth of the actuators. In their current state, the actuators are limited to approximately 1 Hz, whereas the wing’s natural frequency is approximately 13.5 Hz. If the bot is to achieve a higher speed, developers must increase the “flap” frequency to provide the greatest boost. In addition, the researchers identified tethered air supply and non-optimized wing shape to improve future designs.

Despite these current design challenges, the Butterfly Bot has introduced the possibility of increased robotic presence in an aquatic environment. With self-driving cars, parcel-delivering drones, and unmanned satellites, it seems that the ocean has become the final frontier for robotics.