Author: GW Alger

Boston Dynamics released a new video Wednesday, Oct. 30, showcasing the latest capabilities of its humanoid robot, Atlas, which now operates without human oversight. In a simulated industrial environment, Atlas autonomously performs tasks such as moving engine parts between containers and sequencing dollies, using advanced machine learning and sensor technology to detect, grasp and position items. Unlike many robotics demonstrations, Atlas operates without pre-programmed steps or real-time human control. Boston Dynamics says the robot can dynamically respond to changes in its environment, adjusting its approach if, for example, a part doesn’t fit as expected. This adaptability is powered by a combination of camera-based sensors, force sensors and advanced software, enabling Atlas to detect and refine its actions in real time. This level of autonomy contrasts with Tesla’s Optimus robot, which has shown reliance on remote human operators for guidance. In October, Boston Dynamics announced a partnership with the Toyota Research Institute (TRI) to further enhance Atlas’ capabilities with TRI’s behavior models, enabling the robot to quickly adapt to complex tasks required in factory settings. These models, similar to large language models, allow Atlas to learn physical tasks with reduced training needs and improved durability on the job. The newest model of Atlas is fully electric, a shift from its previous hydraulic system, allowing for smoother and more precise movements. Supported by Hyundai, Boston Dynamics sees this collaboration with TRI as a significant step toward developing flexible, humanoid robots capable of performing everyday industrial tasks. CEO Robert Playter described the initiative as focused on creating robots that can handle a wide range of tasks in manufacturing and beyond. Tesla CEO Elon Musk has predicted that by 2040, the global market will include at least 10 billion humanoid robots priced between $20,000 and $25,000. While Atlas marks a leap forward, experts say its long-term success depends on achieving affordability, durability, and scalability to make widespread industrial applications feasible.