Boston Dynamics
Boston Dynamics provided insight on how the company utilizes reinforcement learning to train its Atlas humanoid robot.
Get news, papers, media and research delivered. Sign up for our free newsletters.
Stay up-to-date with news and resources you need to do your job. Research industry trends, compare companies and get weekly market intelligence with Robotics 24/7.
Boston Dynamics
Boston Dynamics provided insight on how the company utilizes reinforcement learning to train its Atlas humanoid robot.
Boston Dynamics provided a behind-the-scenes look into how the company trains its new Atlas humanoid robots.
The new Atlas, which debuted at CES 2026, is a humanoid robot that the company said targets a broad set of capabilities across different facilities, including warehouses or construction sites, that require high levels of strength, endurance and dexterity.
In a video (below), the Boston Dynamics team shared how it is training Atlas to handle tasks like moving objects in a location.
The video features Boston Dynamics’ Alberto Rodriguez, director of robot behavior for Atlas, Shane Rozen-Levy, research engineer, and Vinay Kamidi, research engineer.
The trio wrote about how reinforcement learning and other technical elements are critical to training Atlas how to perform tasks demonstrated in the video, such as moving a refrigerator.
“Atlas uses reinforcement learning (RL) to learn how to lift a fridge by practicing the move with an absurdly large number of variations of the fridge in simulation,” they wrote. “The hardest part is not seeing the fridge or knowing how to lift it, but learning to adapt to whatever version of the fridge that Atlas will encounter in the real world. This is a combined control and perception problem, where perception is done implicitly from body proprioception. The policy driving the behavior has learned to adapt to variations like the location of the fridge, its mass, the amount of grip on the ground and with the fridge, or the configuration where the fridge settles in between the torso, arms, and hands. That level of adaptation is one of the most fundamental building blocks of physical intelligence.”
Click here to read the full article, detailing Atlas’ training.
Artificial Intelligence Deep Learning Machine Vision Machine Learning Components Controllers Motion Control Software Data Management Fleet Management Robot Operating System Simulation News Media Video Press Release Boston Dynamics Humanoid Materials Handling Perception Reinforcement learning
From geometry preparation to AI-assisted analysis, integrated CFD workflows…
GENISOM AI makes ICRA debut at conference in Vienna
North America’s largest robotics and automation event winds down
Automate’s largest day ever draws huge crowds to McCormick Place
