A Robotic Dog that can Go into Unknown Environments and Adapt

Researchers have developed a robotic dog that can take a kick, get back up, and get back on its way

Quadruped robot developers like Boston Dynamics have put hard efforts to develop systems that are capable of traversing all manner of terrain. For the right price, you can now pick up an amazing robotic dog that can take a kick, get back up, and get back on its way.

A team comprised of researchers at Carnegie Mellon and UC Berkeley has created their own system to teach these kinds of robots to make their way over tough ground. The list includes stairs, curbs, and uneven and slippery terrain. Rather than a robotic dog relying on the more standardized method of using cameras to locate the world in front of them, the team trained the roots using simulators: four thousand virtual clones were sent on their way across all manner of different terrain. The researchers claim the method allowed them to effectively reproduce six years of walking experience in a single 24-hour period. After that, the data collected in the simulations was fed into a neural network and then loaded on the robots. 

With onboard learning, the robotic system can react to its environment in real-time and adjust its legs based on that. The team also declared that this robotic system can bring down the cost of robots substantially. "This system uses vision and feedback from the body directly as input to output commands to the robot's motors," researcher Ananye Agarwal mentioned in a post tied to the research. He added "This technique allows the system to be very robust in the real world. If it slips on the stairs, it can recover. It can go into unknown environments and adapt.

Assistant professor Deepak Pathak mentions the system works in similar ways to real animals like cats. "Four-legged animals have a memory that enables their hind legs to track the front legs. Our four-legged robotic dog follows the same way." In addition to being able to climb stairs nearly its own height, the system is also able to work in the dark, although the vision system is yet required for improved performance.