Meta 在其数据中心测试可更换线缆与重置服务器的机器人
Meta Tests Robots That Can Swap Cables And Reset Servers At Its Data Centers

原始链接: https://www.zerohedge.com/markets/meta-tests-robots-can-swap-cables-and-reset-servers-its-data-centers

Meta 正在测试先进机器人技术,以实现数据中心物理任务的自动化,例如更换线缆、重置组件以及服务器维护。虽然该公司声称这些举措旨在解决劳动力短缺并辅助现有员工,但许多员工对工作保障感到担忧,一些人估计机器人最终可能承担高达 80% 的特定技术工作负载。 目前,这些从机械臂到移动平台的机器人仍面临诸多局限;它们通常比人类技术人员速度慢,且难以适应现有基础设施复杂的以人为本的设计。尽管存在这些障碍,Meta 仍将机器人技术视为在人工智能密集型环境中提高效率的长期战略,机器人可以在不适合人类工作的条件下运行。 除了运营方面的担忧,这一转变还引发了关于数据中心劳动力未来,以及基于创造就业机会的政府税收激励措施有效性的质疑。虽然人类在灵活性和适应性方面目前仍具优势,但 Meta 对自动化的持续投资预示着未来机器将承担数据中心大部分的体力劳动,这可能会迫使行业角色和地方经济政策发生转变。

相关文章

原文

Kiss those data center technician jobs goodbye...

Meta is testing robots that can swap network cables, restart servers, reseat components, and perform other physical tasks traditionally handled by data center technicians. The experiments come as the company rapidly expands its AI infrastructure and seeks to operate facilities more efficiently, according to Wired.

Some employees worry about job security. One Meta worker estimated that a successful cable swapping robot could eventually take over as much as 80 percent of certain workloads. “We thought those of us performing the physical tasks were safe for a while, but not anymore,” the worker said.

Meta says automation does not mean it needs fewer people. Spokesperson Francis Brennan cited a shortage of skilled tradespeople and the company’s investments in training and hiring workers. Meta has launched programs teaching electrical, mechanical, and plumbing skills and partnered with trade unions on apprenticeships.

Still, robotics is part of Meta’s longer term strategy. Robotics manager Eric Xu has said robots could eventually assist with incident response, environmental monitoring, and preventive maintenance.

The company is experimenting with equipment from several manufacturers. A Kinova Gen3 arm is being evaluated for power cycling servers, while other machines are being tested for cable replacement. Some facilities already use a simple remotely controlled device resembling a mechanical finger to press power buttons and reboot equipment.

At its Altoona, Iowa, campus, Meta is testing dual arm Watney robots for cabling. At its newer Prometheus campus in New Albany, Ohio, ABB robots mounted on four wheel platforms are being used to reseat components and could eventually perform more work with less human supervision.

These projects build on simpler automation already deployed in Meta facilities. Self driving tugger robots transport heavy server racks, while wheeled inventory robots scan equipment and assist with inspections. Microsoft, Google, and Amazon have also explored robotics for data center operations.

Wired writes that the economics are attractive. Robots could provide consistent labor where qualified technicians are scarce and handle repetitive or hazardous work. They could also operate in hotter, darker, or otherwise less hospitable environments.

But current systems remain far from replacing technicians entirely. Meta’s inventory robots struggle with cables and corners, require humans to move them between buildings, and cannot reliably interpret some equipment indicators. Other robots need substantial charging time and remain slower than people.

Data centers were also designed around human dexterity. Complex cabling, particularly around advanced AI systems such as Nvidia’s GB300, remains difficult for robots. As one former Meta employee put it, “Things have been designed for human hands forever to make everything a five-minute repair.”

Even so, the experiments are changing how some Meta workers view their future. Employees have reportedly discussed fears that automation could eliminate jobs or shift remaining positions toward lower paid workers who mainly follow AI generated instructions.

That could also affect the politics surrounding data centers. Communities often justify tax incentives partly through the jobs these facilities create. If robotics significantly reduces employment, governments may reconsider those economic tradeoffs.

For now, humans remain faster and more adaptable. But as robotic hardware gets cheaper and AI improves, Meta is preparing for a future where machines perform considerably more of the physical work inside its data centers.

联系我们 contact @ memedata.com