These workers help train robots who could one day steal their jobs

Inside a textile plant in Karur, southern India, employees have been asked to attach cameras to themselves while they work. The devices are intended to record the fine movements behind routine production tasks, creating material that could be used to teach artificial-intelligence systems how to perform them. For workers, the request places a familiar factory job at the center of a larger question: whether the information collected from people today could eventually support machines that need fewer people tomorrow.

On a recent day at Texworth, staff distributing GoPro cameras and smart glasses moved through the expansive production area. One of the people asked to wear a camera was Shoba Kirubakaran, 40, who was putting yellow, chicken-patterned aprons into plastic packaging for the Designs By Danica brand. Kirubakaran said this was the third occasion on which she had been instructed to use a GoPro during her shift.

The camera records the sequence of her hand movements as she packages the goods. That sort of footage can give developers examples of how a human approaches an object, positions it, completes a task and moves to the next one. In this case, the aim is to provide training material for AI-powered robots that may be designed to carry out the same work. The collection process turns everyday manual labor into data as well as output.

Kirubakaran said she had not been told that the recordings were connected to robot training before NPR raised the subject with her. “Nobody told me anything about it,” she said. The gap illustrates a central tension around workplace data gathering: an employer may see video as part of a technology project, while an employee may understand it simply as another instruction to follow on the factory floor.

For Kirubakaran, the stakes are personal. She has spent a decade at the plant, working six days each week, generally from about 9 a.m. until 6 p.m. Her monthly pay is $90. Together with her husband, who drives a rickshaw, she has used that income to support their two sons. She said her experience is tied to this particular line of work: “I only know how to do this job.”

The experiment reflects a broader push to move robots beyond highly structured tasks. Factory automation has long been most practical where motions are repeated and the setting changes little. Packaging can look simple, but it demands coordination with soft materials, plastic wrapping and products that must be handled in sequence. Video captured from a worker’s point of view may offer a detailed record of those choices, including motions that are hard to describe in written instructions.

But recording a task is not the same as proving a robot can replace the person doing it. A useful system would still have to translate footage into dependable physical action and continue working safely when materials, positioning or conditions vary. The available account does not establish whether robots trained with this footage can perform Kirubakaran’s assignment, when they might be deployed, or how many jobs could be affected. Those unanswered questions matter as much as the cameras themselves.

There is also a different way to view the process. Supporters of automation often argue that technology can take on repetitive labor, while critics focus on who carries the risk when lower-paid workers help create the knowledge that could reduce demand for their roles. At Texworth, the immediate result is not a robot on Kirubakaran’s line, but a record of how she does the job. Whether that record becomes a working machine, and what protections or alternatives workers receive if it does, remain the issues to watch.

For now, the project makes the future of automation unusually visible: it is mounted on workers’ bodies during an ordinary shift. Kirubakaran’s actions packaging aprons are being preserved not only as part of a day’s production, but potentially as lessons for software and machines. Her experience underscores how advances in robotics can begin quietly, through the observation of the people whose skills make factories run.

Source: NPR Health