Apr 25 2013
Dubbed “Flipperbot”, the robot has been presented today, 24 April, in IOP Publishing’s journal Bioinspiration and Biomimetics, and was designed to test how real-life organisms, such as seals, sea turtles and mudskippers use flippers and fins to move on surfaces such as sand.
Flipperbot. Credit: Nicole Mazouchova
The 19 cm-long robot was built by Nicole Mazouchova, working in Professor Daniel Goldman’s Complex Rheology and Biomechanics (CRAB) Lab at the Georgia Institute of Technology, and Dr Paul Umbanhowar from Northwestern University.
It weighs just 790 g and is not propelled by legs or wheels but instead crawls using two flipper-like front limbs, spanning 40 cm. Each limb is driven by small servo motors and has a thin, lightweight flipper at its end. During the study the robot was tested on a 122 cm-long bed of poppy seeds and was recorded using a high-speed digital camera.
A video of Flipperbot in action can be viewed here: https://www.youtube.com/watch?v=s0_elE74Mdc.
The researchers believe that the improved understanding of flipper-based locomotion gained from their study could inspire the design of future multi-terrain robots that can swim and walk effectively using the same appendages.
Furthermore, the findings from this study could possibly be used to shed light on the evolutionary adaptations of structures such as fins or flippers when fish-like animals moved from aquatic to terrestrial environments several hundred million years ago.
Professor Daniel Goldman and his group have previously studied the high walking performance of hatchling sea turtles, and the goal of this new study was to use a robot to delve deeper into the mechanics of flipper-based movement on land.
“Flipperbot allowed us to explore aspects of the sea turtle’s gait and structure that were challenging, if not impossible, to investigate in field experiments using actual animals,” said Professor Goldman.
The research has shown, among other things, that a free wrist at the end of the flipper holds a significant advantage over a fixed wrist.
“One of the main findings of our paper was that when the robot was fitted with a free wrist, it was able to move more effectively over the ground as it allowed the flipper to remain locked in place within a solid region of sand and thus disturbed less material during the forward thrust.
“With a fixed wrist, the robot also interacts with the ground that has already been disturbed by its previous steps, which hinders its movement,” continued Professor Goldman.
These conclusions were backed up by real-life footage of hatchling loggerhead sea turtles taken by study co-author, Nicole Mazouchova, during a six week field-study on Jekyll Island, Georgia, which showed that the sea turtles experienced similar failures.
Mazouchova believes that further robot testing could help in turtle conservation biology. She said: “The natural beach habitat of hatchling sea turtles is endangered by human activity. Robot modelling can provide us with a tool to test environmental characteristics of the beach and implement efforts for conservation.”
As well as the wrist flexibility, the researchers also investigated the different depths to which the flipper penetrated the poppy seeds, while also measuring the body lift, flipper thrust, the drag of the robot’s base and the amount of ground that was disturbed. The results showed that the relationship between each of these aspects is not trivial, but can, nevertheless, be understood using relatively simple models.
Co-author of the study Dr Paul Umbanhowar said: “Our modelling shows that flipper driven locomotion on soft ground is largely determined by simple granular physics. But the inherent sensitivity of this type of movement to disturbed ground means that very small changes in gait or body structure can cause dramatic decreases in speed.”