RoboBusiness talk to highlight teaching robots through human actions
At RoboBusiness, OLogic CEO Ted Larson will present on egocentric robot learning, a technique that trains robots using first-person human demonstrations captured via wearable cameras and teleoperation. This method aims to cut the time and cost of deploying robots in sectors such as manufacturing and logistics. The session will address technical hurdles and the future of commercial robotics.
Egocentric robot learning relies on first-person video and teleoperation data to train machines in manipulation and decision-making. The approach has drawn backing from major AI research organizations, and OLogic's project portfolio spans Bear Robotics' autonomous mobile robots, Locus Robotics' warehouse systems, and social robots from Zoetic and Misty Robotics, illustrating its commercial range.
Larson brings over two decades of robotics and consumer electronics design experience, having collaborated with companies such as Google, Hasbro, and Panasonic. His RoboBusiness session will address technical obstacles and how embodied AI and multimodal data collection could accelerate robot deployment across manufacturing, logistics, and healthcare.
Egocentric robot learning could lower automation barriers by letting companies train robots without deep programming expertise. Smaller manufacturers and logistics firms may adopt robotics more readily, potentially easing labor shortages while reshaping job roles. Workers might shift from manual tasks to supervisory or data-collection positions, though the transition could create uncertainty for those whose skills become automated. The approach's success depends on reliable data collection and cost-effective deployment.