Most of the robots created by engineers and scientists around the world are designed to make our lives much better, and many of them to help us in rescue tasks in the face of major natural disasters.
However, robotics has also improved to the point of creating robots just for leisure, and surely you’ve heard about the famous RobotCup, one of those sporting events just for robots and now they have a new member.
By name DribbleBot, It has been created by the Massachusetts Institute of Technology (MIT), a quadrupedal robotic dog that is able to navigate gravel, grass, sand, snow, and pavement well, and can easily pick itself up if it falls.
Specifically, it has been created by researchers from the Computer Science and Artificial Intelligence Laboratory (CSAIL) and its movements are possible thanks to the integration of new technologies such as machine learning, integrated sensors, actuators, cameras and computing power.
However, for the robot to be able to carry out these soccer movements, it has to have previously learned them. To do this, the researchers have strengthened the classic method of positive reinforcement when the robot performs a successful movement.
“One aspect of this reinforcement learning approach is that we need to design a good reward to make it easier for the robot to learn successful dribbling behavior.“, says Gabe Margoliswho co-led the work with Yandong Ji, a research assistant at AI Lab.
“Once we have designed that reward, then it is practice time for the robot: in real time, it is a couple of days, and in the simulator, hundreds of days. Over time, he learns to get better and better at handling the soccer ball.“, Add.
With this, the robot takes into account the friction and drag interactions of a ball with the ground and observes how it differs when facing different terrains.
Not only for football, but also for rescue work
But this robot It has not been created just to play football, but to move on terrain where human beings, or others robotsthey cannot do it, and thus use it in rescue tasks.
“If you look around you today, most robots have wheels. But imagine that there is a disaster scenario, floods or an earthquake. We want robots to help humans in the search and rescue process“, it states Pulkit AgrawalCSAIL Principal Investigator.
“Our goal in developing algorithms for legged robots is to provide autonomy in challenging and complex terrain that is currently beyond the reach of robotic systems.”, he adds.
