Biomimetic Backflips: When Robots inform Biology
Posted by Jenna Capyk on November 2, 2011
The area of biomimetics is pretty much what it sounds like: mimicking biology. These principals are used extensively for many types of engineering problems and are now an area of very prolific research in robotics. It’s also a really logical way to proceed, considering the mechanism of = natural selection. Through natural biological evolution, the physical traits best suited to perform a specific task develop and are then refined. This means that many the traits we see today have had hundreds of thousands of years of improvements, and are impeccably suited for their specific role. By borrowing the physical strategies that we observe in nature, we are basically hijacking millennia of evolutionary innovation. Luckily, Nature has notoriously inept patent lawyers.
A lot of the time, the goal of biomimetic research is to find out how to make the stuff we want to create better by using strategies that we find in nature. Insects are often used as biomimetic models because their basic body plan and exoskeleton make their traits particularly amenable to robotic copying. This is the case for the robot “Dash,” modeled off of the cuddly cockroach. It is designed as a tough, all-terrain scampering robot that can go where it may be too cramped or dangerous for humans to go. What the researchers who created Dash maybe didn’t expect, however, was for this metal critter to provide some real insight into evolutionary biology.
To solve some of the balance problems that Dash was having in it’s clambering activities, the researchers added flapping wings and a tail fin from a toy airplane. Interestingly, this helped the running ability dramatically. The degree of the slope the robot was able to handle was greatly increased, both up and down, and the speed of the electronic creeper’s scuttle almost doubled.
Although these improvements were impressive in and of themselves, they were also a test of some theories in the evolution of flight that before had only been supported by a scant fossil record. In this field, computer models had suggested that for biological scamperers, they would need to triple their running speed to be able to achieve flight. Dash’s wings doubled the speed, but fell short of this mark. The experiment therefore actually provides evidence for a theory of flight evolution suggesting that flight started as gliding in tree-dwelling creatures, rather than from land-based animals requiring “takeoff speed”. Obviously it’s not conclusive as there are a lot of variables, but an interesting and partially accidental experiment none-the-less.
It’s interesting to think that while we use nature as a model to create things in science, the very things we create can be used as a model to help understand nature. Ah science, you never know which way it’s going to go!