Robotics & Automation News Publishes Comparison Guide on Warehouse Automation

On November 3, 2026, Robotics & Automation News published an article on choosing between humanoids and autonomous mobile robots (AMRs) for warehouses.

The article is a comparison guide to warehouse automation options, not an announcement of a specific robot deployment or performance test. The focus when reading it as news, therefore, is less on the publication of the comparison itself than on how much evidence is available to inform decisions about real-world warehouse operations. See the original article for details. In cases we have reported, Agility Robotics announced plans to expand Digit deployments at two GXO sites in the United States (GXO and Agility Robotics Expand Digit Deployments at Two U.S. Sites). The company’s RoboFab manufacturing facility in Salem is designed for an annual production capacity of up to 10,000 units ([Agility Robotics Announces Growing Enterprise Demand for Digit and Operates Manufacturing Facility With Annual Capacity of 10,000 Units](/en/news/agility-robotics-announces-growing-enterprise-demand-for-digit-operate-3a7e60)). Agility Robotics also announced that it would standardize Digit 5’s onboard computing platform on NVIDIA IGX Thor (Agility Robotics Standardizes Digit 5’s Onboard Computing Platform on NVIDIA IGX Thor). These provide information on different stages—deployment sites, manufacturing capacity and the robot’s computing platform—and illustrate why warehouse automation cannot be assessed solely by the capabilities of a single robot. However, deployment at sites and plans for annual production capacity alone do not show which approach is suited to which tasks or whether it offers operational advantages. Summaries of previously reported information do not provide figures comparing utilization, throughput or costs including maintenance for humanoids and AMRs performing the same tasks. Data comparing the two under consistent task and deployment conditions, as well as operational results and records of downtime and maintenance, will need to be examined. Without these, individual deployment expansions cannot establish the superiority of one approach over another.