'Terradynamics' Predicts Robot Mobility on Granular Surfaces

New research from Georgia Tech could help create and advance the field of 'terradynamics' - a name the researchers have given to the science of legged animals and vehicles moving on granular and other complex surfaces. Predicting this type of movement could allow designers to optimize legged robots operating in complex environments for search-and-rescue missions, space exploration, or other tasks. The Georgia Tech researchers examined the motion of a small legged robot as it moved on granular media. Using a 3D printer, they created legs in a variety of shapes and used them to study how different configurations affected the robot's speed along a track bed. They then measured granular force laws from experiments to predict forces on legs, and created simulation to predict the robot's motion.

Transcript

00:00:02 here in the school of physics at Georgia Tech we're interested in how organisms manage to move around in complex environments one component that allows animals to move so well is that they have Limbs and legs which can be multifunctional they can allow the animal to climb over Ledges they can allow the animals to Sprint rapidly over hard ground to paddle through soft

00:00:23 ground and even potentially kick through fluids so we're interested in exploring how uh limbs can enable robots to move around with the agility and Mobility that some animals with limbs have the idea is to begin to develop a terod dnamic equivalent to arrow and hydram which will allow us to predict mobility of devices in these complex environments the robot that we created uh is based on

00:00:47 a simple toy robot and then we took its legs off and replaced those with legs that we 3D printed and we found that legs with a positive curvature such that the open part of a c is pointing upward uh worked on average better than legs with a negative curvature producing more thrust Force as well as more lift Force we took these Force relationships that we measured and we input them into a

00:01:10 special kind of simulation called a multi-body dynamic simulation we found that when we put the force relationships into the simulation we could predict the robot speed to a few percent over a pretty wide range of conditions of leg geometry and and granulate material we think that for roboticists and Engineers who are interested to design vehicles that can move on loose

00:01:35 granular material they can immediately begin to optimize limb shapes and limb trajectories of small robots in a real search and rescue robot the type of terrain that this robot will encounter is not even necessarily granular material but we think that one component could be the type of loose material which robots are styed on and which animals can often navigate so well and

00:02:00 we think that this kind of terod dnamic of Granny material will allow design of devices with better performance than what's currently available