Huge developments were made this past Wednesday that could mean big things for disaster relief operations.
Scientists have invented a small robot that has adaptive capabilities and is able to bounce back from damage to its physical structure. The function of these robots is to use them as first-responders. Robots that today would be put out of commission by certain physical setbacks can now stay operational with a slight change in performance.
Jean-Baptiste Mouret says that the aim “is to have robots that can survive in hostile environments such as a Fukushima-type nuclear disaster. If we send in robots, they have to be able to pursue their mission even if they are damaged, and not just come to a halt in the middle of a reactor.”
These amazing machines were inspired by how learning naturally occurs. We learn via trial and error. If we have an injury we instinctively go through a trial and error process until we find something that allows us to continue functioning in a way that won’t hurt us further. If I injure my thumb, for instance, I will need to find a way to continue what I was doing without using it. So I can try to use my other fingers or, if need be, use my other hand. Experience gives us this instinctive, natural problem solving that scientists have attempted to replicate in the programming for these robots.
The programming is essentially based on the movements the robot is capable of. Each movement has a “value” which is assigned based on its usefulness in an emergency situation. These “values” guide the robots learning and problem solving patterns. This algorithm is called Intelligent Trial and Error. The robot learns to reject or employ certain movements and behaviours based on their “values”.
So what do they look like? They are walking, 20 inches and have six legs.
In trials, a robot that had one third of its legs broken was able to find a way to continue on despite the damage. Antoine Cully says that the robot tries one action but if it “doesn’t work, the robot is smart enough to rule out that entire type of behaviour and try a new type. For example, if walking mostly on its hind legs does not work well, it will next try waking mostly on its front legs. What’s surprising is how quickly it can learn a new way to walk.”
“It’s amazing to watch a robot go from crippled and flailing around to efficiently limping away in about two minutes.” – Antoine Cully