Researchers and engineers have begun to unravel the mysteries of evolution. Yes, we know that DNA changes and favorable mutations are awarded with longer lives and more offspring, but how can scientists tell which evolutionary step was really the best? Computer programs can help to simulate different scenarios but, as many scientists will attest, there is no substitute for the real thing. That is where John Long of Vassar College steps up to the plate. He uses robotics to mimic the movements of species and can change their characteristics ever so slightly to determine exactly what a specific evolution does. These robots can tackle the tough evolutionary questions, not just why the dodo didn’t make it.
Scientists and researchers have brought computer simulation forward by leaps and bounds within the past few years. Take a look at the sport of Formula 1 and it will be evident. Every single part of the car is modeled using hundreds of hours of computer simulations and wind tunnel testing. Despite the extensive computer work, engineers still find that the simulations miss a few key points that are only resolved with track testing. There are little niggles in fluids simulations that are just not quite perfect and could make or break any type of testing, not just Formula 1. It is in the real world that the best of testing is still being conducted and Dr. Long knows that.
The most high-profile case of the type of research occurred in Dr. Long’s lab, where he was asked to unravel the mystery of the newly discovered Predator X. The pliosaur, named Predator X for its purpose-built predator nature, was a 45-ton killing machine that owned the oceans in the Jurassic Period. Researchers, however, were unclear as to why the pliosaur had two sets of symmetrical flippers. Most modern reptiles also have two sets of flippers, but they are not symmetrical and each has different a purpose: either propulsion or steering. The symmetry of the flippers on Predator X indicates that all four flippers were used for propulsion.
Dr. Long then fitted a swimming robot named Madeleine with the symmetrical flippers and did a few tests. He found that, though a four-flipper swimming reptile had a low fuel economy (considered the Hummer of the ocean), its capabilities for acceleration were much greater than the two-flipper counterpart. Predator X might have had a heavy appetite, but it was more than capable of catching all but the speediest of prey.
Another stallion in Dr. Long’s stable is Preyro, a robot capable of playing predator/prey that is easily modified to test certain evolutionary possibilities. The evolutionary change that results in Preyro not being caught display the exact mechanics of natural selection. The only difference is that engineers no longer need to wait for a specific evolution to materialize, they can create one within the robot and test it out. For Preyro, Dr. Long changed the number of vertebrae that covered its mechanical tail. The added stiffness would change the speed at which the robot could swim away from the predator. His studies found that exactly 7 vertebrae allowed Preyro to swim the fastest, with fewer or greater numbers giving a slower getaway. This type of research gave an instant answer as to why fish tails evolved as they did. Take a look at the video for more information on Preyro.
The research conducted over robotic evolution is focused mainly on the past, so what bearing could it possibly have on the future? As current robots get better at mimicking past and present real-life situations, the researcher’s knowledge base will continue to grow. The extra bit of information gained in these experiments could not only give engineers a few more options in robot design, but will also serve as good practice in the production of new and unique robots. When a robot is needed for a future application, there will be less need for extensive testing, making it possible to create new robots faster and more efficiently. As they always say, practice makes perfect.