Robotics and Embodied AI
Integration of Intelligence with Physical Form
A timeline of key developments in robotics and embodied AI — from early industrial systems and mobile perception to modern humanoid platforms integrating large language models.
Unimate: First Programmable Industrial Robot
George Devol and Joseph Engelberger develop Unimate. Devol files the original patent in 1954 for a “Programmed Article Transfer” device. In 1961, the first unit begins operation on a General Motors assembly line at the Inland Fisher Guide plant in Ewing Township, New Jersey, performing die-casting handling and spot welding.
Shakey the Robot
Shakey the Robot at the Stanford Research Institute demonstrates early mobile perception and navigation. Shakey integrates a television camera, range finder, and bump sensors with a reasoning program, becoming one of the first mobile robots to interpret instructions, plan actions, and navigate through a real-world environment.
iRobot Roomba
iRobot releases the Roomba, bringing autonomous navigation to consumer vacuum cleaners. The device uses infrared sensors, bump sensors, and algorithmic cleaning patterns to navigate domestic spaces, becoming one of the first commercially successful autonomous consumer robots.
Boston Dynamics: Atlas and Spot
Boston Dynamics develops dynamic platforms such as Atlas and Spot, advancing legged locomotion. Atlas demonstrates bipedal walking, jumping, and backflips, while Spot — a quadruped — demonstrates stable navigation over uneven terrain, stairs, and construction sites. These platforms showcase advances in dynamic balance, sensor fusion, and real-time motion planning.
Foundation Models Meet Robotics
Integration of large language models and multimodal AI improves instruction handling and visual processing in robots. Pairing foundation models with robotic systems allows robots to interpret natural language commands, identify objects in unstructured environments, and adapt to novel tasks without explicit programming for each scenario.
Humanoid Platforms Emerge
Companies demonstrate humanoid robots with enhanced dexterity and general-purpose capabilities. Tesla reveals Optimus, a humanoid platform intended for repetitive and dangerous tasks. Figure AI demonstrates its Figure 01 and Figure 02 robots performing warehouse and logistics tasks. Unitree introduces the G1, a compact humanoid attracting attention for its agility and accessible price point.
Commercial Deployment Advances
Production versions advance toward commercial deployment. Boston Dynamics unveils its fully electric Atlas at CES 2026 with increased degrees of freedom for industrial applications. Unitree’s G1 gains attention for demonstrations of dynamic movement including running, jumping, and recovery from pushes. Pilot programs expand in manufacturing and logistics settings.
Persistent Challenges
Reliable operation in unstructured environments remains difficult. Energy efficiency continues to limit deployment duration. Safe human-robot interaction requires advances in sensing and behavioral prediction. The “sim-to-real gap” — transferring behaviors learned in simulation to physical robots — remains an active area of research.
This page is maintained as a living reference and will be updated periodically as new developments occur.