Humanoids Deployed on Factory Floors

The humanoid robot BMW is deploying is Apptronik's Apollo, a 5'8" tall, 160-pound machine capable of lifting 55 pounds. It's designed to work in spaces made for humans and can run for 4 hours on a battery pack, with hot-swappable batteries enabling up to 22 hours of daily operation. This deployment is an expansion of a partnership that began with Figure AI's 02 robot at BMW's Spartanburg, South Carolina plant in 2025. The Apptronik team has deep roots in robotics, with key members having worked on NASA's Valkyrie humanoid robot at the University of Texas at Austin. This expertise has attracted significant investment, with Apptronik raising over $935 million in total Series A funding, including a recent $520 million extension with backing from Google, Mercedes-Benz, and John Deere. The company is also collaborating with NVIDIA, using its Project GR00T AI foundation model to enable Apollo to learn tasks from human demonstrations. Xiaomi's CyberOne humanoid, standing 5'9" and weighing 115 pounds, is being integrated into the company's electric vehicle manufacturing. CEO Lei Jun has stated that he expects humanoids to be working at a "large scale" in Xiaomi factories within the next five years. The robot can perceive its environment in 3D and recognize 85 types of environmental sounds and 45 classifications of human emotions. The Foxconn-Nvidia partnership in Houston is centered on creating a "benchmark AI smart factory." They are leveraging Nvidia's Omniverse for factory simulation and the Isaac platform for robot training and path planning, aiming to deploy humanoids powered by the Isaac GR00T model in the first quarter of 2026. This "simulation-first" approach is intended to overcome the slow and expensive process of collecting real-world robot data. These deployments are part of a broader push to automate tasks that are difficult, unsafe, or tedious for human workers. While single-purpose robotics have been used for decades, the goal of general-purpose humanoids is to provide greater flexibility in environments already designed for people. This is seen as a key step in addressing labor shortages and improving operational efficiency in manufacturing and logistics. However, the transition from controlled demos to widespread, reliable deployment faces significant hurdles. Key challenges include ensuring sustained uptime, improving dexterity for fine-motor tasks, and radically reducing costs to make the technology economically viable at scale. Industry analysts note that while intelligence and perception capabilities are advancing rapidly, battery life and handling precision remain significant gating factors for broader adoption.