Meta is actively testing robotics to automate physical tasks inside its data centers, according to current and former workers. The social media giant is utilizing hardware from vendors including Watney Robotics, Kinova, and ABB to test capabilities such as swapping networking cables, power cycling servers, and performing basic hardware resets. One worker estimated that successful deployment could automate up to 80 percent of certain physical workloads.
While Meta declined to comment directly on the specific testing, company spokesperson Francis Brennan emphasized ongoing investments in human hiring amid a national skilled labor shortage. However, internal statements from Meta robotics leadership confirm long-term goals to deploy robots for incident response, environmental monitoring, and preventative maintenance across its expanding infrastructure footprint.
The push highlights how hyperscalers are turning to hardware automation to control soaring operational costs as AI compute demands scale. Advances in AI models powering physical robotics have reduced server damage risks and improved precision, paving the way for uncrewed data center operations in remote or extreme environments.
Why it matters
Hyperscalers are seeking to lower data center O&M costs through physical automation as AI infrastructure spending skyrockets.
Robotics startups targeting industrial and data center tasks face immediate commercial opportunities with large cloud providers.
Data center operations are shifting toward higher automation density, impacting future facility design and labor requirements.
Source: arstechnica.com
