Here’s a fun challenge: teach a quadruped robot to successfully dribble a soccer ball. It is essentially a core component of RoboCup, the huge international competition founded in 1996. Soccer is a great way to test a robot’s locomotion, agility, and decision-making.
Two key differences from MIT’s Dribblebot: First, RoboCup robots are generally bipedal. Second, and more important, this robot is designed to perform complex tasks on uneven and changing terrain, which adds another level of difficulty to the task.
“The above approaches simplify the dribble problem, making a hard, flat terrain model assumption,” project co-leader Yandong Ji said in a post linked to the news. “The movement is also designed to be more static; the robot is not trying to run and handle the ball simultaneously. That is where the most difficult dynamics come into the problem of control. We address this by extending recent advances that have allowed for better locomotion outdoors in this composite task that combines aspects of locomotion and dexterous manipulation together.”
Potential terrains include grass (naturally), sand, gravel, mud, and snow. The answer to all of the above is one that should be familiar to anyone familiar with the robotics space of late: simulation, simulation, simulation. In training, the physical robot is considered a “digital twin,” put to the test as the computers run 4,000 simultaneous simulations of different environments.
This type of training clearly has broader applications beyond the limited world of robotic soccer. The debate over the effectiveness of legged robots continues, but one thing is certain: there are limitations to how far you can currently go on wheels.
“If you look around you today, most robots have wheels. But imagine there is a disaster scenario, a flood or an earthquake, and we want robots to help humans in the search and rescue process. We need the machines to traverse terrain that is not flat, and wheeled robots cannot traverse such landscapes,” says MIT professor Pulkit Agrawal. The goal of studying legged robots is to go into terrain beyond the reach of current robotic systems.”
Of course, Dribblebot also has its own limitations. Stairs and slopes are still a challenge for the little robot.