The short version
- Meta is conducting internal tests with various robotic systems to perform physical tasks such as plugging in cables and resetting servers within its data centers.
- Company leadership argues that automation addresses a critical shortage of skilled labor, while workers express fear that their roles are becoming obsolete.
- The shift toward robotics could allow for operations in extreme environments and may lead to a centralization of technical oversight in fewer locations.
Meta is actively testing robotic systems designed to perform physical maintenance tasks inside its data centers, marking a significant step toward automating roles traditionally held by human technicians. According to current and former employees familiar with the initiatives, these trials involve hardware capable of plugging in networking cables, resetting servers, and handling other routine operations. The effort remains largely unpublicized but represents a strategic move to manage labor costs as the company’s spending on artificial intelligence infrastructure continues to rise rapidly.
The experiments utilize equipment from multiple vendors, including Watney Robotics, Kinova, and ABB. In one specific trial, engineers are evaluating whether a Kinova Gen3 robotic arm can effectively power cycle servers by cutting off and restoring electricity. Another test focuses on a different robot designed to swap networking cables. At some facilities, the company has quietly deployed simpler devices that resemble fingers or pointed sticks. These tools press power buttons on devices like Mac Minis to restart them when prompted remotely by human operators.
Workers involved in these projects have expressed significant anxiety regarding their job security. One data center employee estimated that if the robotic trials prove successful, the machines could replace up to eighty percent of certain workloads. The sentiment among staff is one of resignation, with some noting that physical labor roles previously considered safe from automation are now under threat. Internal communications within worker groups have reportedly taken on a demoralized tone, with discussions centered on the likelihood of widespread displacement in the coming years.
Meta’s official stance contrasts sharply with these internal concerns. A company spokesperson emphasized that Meta is heavily investing in hiring and training workers to build and operate its facilities. The statement highlighted a broader national infrastructure boom and cited a major shortage of skilled labor, arguing that the industry needs more workers rather than fewer. This position aligns with public relations efforts that often tout job creation as a benefit of data center expansion, which has been used to justify tax breaks in communities hosting these facilities.
However, internal leadership has outlined different long-term objectives. Eric Xu, a senior manager for robotics at Meta, stated last year that the company’s goal is to integrate robots into data centers to speed up incident response and handle preventative maintenance. He argued that starting these initiatives now is essential for future success, even if it requires more collaboration and research. This perspective suggests that while immediate hiring needs exist, the strategic direction points toward reducing human intervention in routine tasks.
The feasibility of such automation has improved due to advancements in hardware and artificial intelligence. Earlier industry trials faced challenges because robots were expensive and prone to damaging delicate server equipment during simple tasks. Recent developments have lowered costs and enhanced the capabilities of AI models powering these machines. Several startups are now developing commercial robots capable of assembling servers, swapping parts, and replacing wires with greater precision than previous generations.
Proponents of this technology argue that robots can provide consistent output at a lower cost, particularly in regions where data centers are abundant but qualified workers are scarce. There is also an argument that automation can improve working conditions by taking over repetitive and dangerous tasks, allowing human technicians to focus on more complex challenges. In the long term, reducing reliance on human presence could enable data centers to operate in extreme environments, such as underwater or in space, where conditions are inhospitable to living beings.
The implications extend beyond operational efficiency to labor dynamics and community relations. In Altoona, Iowa, where Meta operates its largest data center campus, the company benefits from significant property tax breaks tied to job creation. If automation reduces the need for local technical staff, it could undermine the economic justification for these incentives. Additionally, there are concerns that AI-driven management tools may allow Meta to centralize certain roles in locations like Denver, Colorado, potentially leading to a decline in the quality and pay of remaining jobs in other regions.
Sources behind this briefing
Go to the original reporting
- Ars Technica↗Inside Meta’s push to put robots to work in data centers
