Researchers at the University of Southern California developed a robotic foot that teaches to try and try, similar to an animal.
The team, of the USC Viterbi School of Engineering, believe that they are the world's first to create a leg operated by veins like tendons that can even fix their foot after they get tangled up, so it does not fall.
The algorithm working on the leg can learn a new task of walking after only five minutes of "play" – by making random movements that create a map of the limb and its interactions with its surroundings.
"Nowadays, it takes the equivalent of months or years of training for a robot to be ready to communicate with the world, but we want to achieve quick learning and adjustments that are seen in nature," said Professor Francisco de Valero-Cuves, who worked on the project with Doctoral students Ali Marianedegad, Dario Urbina-Melendez and Brian Con.
Going for a walk
Researchers hope technology can have several possible uses, including our understanding of human movement and disability and the creation of robots that can move through difficult and rapidly changing environments (for tasks such as disasters).
"If you release these [new] robots learn from relevant experience, then they will eventually find a solution that, once found, will be put to use and adapted as needed, "said Marianidežad, the lead author of the research book.
"The solution may not be perfect, but it will be adopted if it is good enough for the situation. Not everyone needs or wants – or can spend time and effort – to win an Olympic medal."