The New Gig: Teaching Robots to Fold Laundry
While Silicon Valley and Shenzhen chase headlines with flashy humanoid demos, a quieter — and arguably more important — robotics story is unfolding in Indian homes, garment factories, and rented office spaces across the country. Over the past year, a wave of Indian startups has emerged with one job: recruit ordinary people to record themselves doing everyday tasks, and sell that data to the robotics companies racing to build the next generation of AI-powered machines.
The workers aren’t engineers. They’re schoolteachers moonlighting after class hours, homemakers strapping smartphones to their heads while washing dishes, and garment factory employees going about their normal shifts — all while cameras capture every hand movement, every grip, every micro-adjustment. This first-person, or “egocentric,” video is exactly what robotics companies need to train the vision-language-action (VLA) models that let humanoid robots understand and replicate human movement in the real world.
Why India, and Why Now
The logic driving this boom is straightforward. Building an AI-enabled humanoid robot isn’t just a hardware problem — it’s a data problem. Robots trained purely in simulation struggle when they hit the messiness of real kitchens, warehouses, and factory floors. Companies need enormous volumes of real-world human movement data to close that gap, and that data has to be collected, verified, and organized at scale.
India offers a combination that’s hard to find elsewhere: the world’s second-largest workforce, English-language capability, and comparatively low labor costs. Data-collection firms recruiting Indian workers say most of their contracts come from companies based in the US and China — either directly from robotics manufacturers or from intermediary firms that specialize in sourcing and annotating this kind of training data.
This isn’t a small, informal side hustle anymore. North India-based startup Neocambrian AI recently opened a dedicated robotics data factory in Noida, built specifically to capture robotics data in simulated environments — and has built out a network of over 100 partner sites where workers record themselves performing everyday tasks. Multiple similar operations have sprung up around the country in under a year.
A Market Racing Toward Trillions
The scale of investment behind this trend is significant. Global brokerages have issued eye-catching forecasts for the humanoid robot market — Barclays projects the sector reaching $200 billion within a decade, while Morgan Stanley has forecast the market surpassing $5 trillion by 2050, with the global robot population reaching 1 billion units. Every one of those robots will need training data drawn from real human behavior, and that’s where India’s labor pool becomes strategically relevant — not as a robot manufacturing hub yet, but as the data backbone for robots built elsewhere.
The Catch: A Race to the Bottom on Price
For all the opportunity, the economics are already shifting under the feet of the people doing the work. Industry insiders report that contract prices for data collection have already halved within a matter of months, as more firms enter the space and compete for the same client contracts. Data collection, in other words, is being commoditized fast.
That’s forcing a reckoning for India’s emerging robotics-data ecosystem. Experts tracking the space argue that Indian firms currently positioned purely as data-collection vendors will need to move up the value chain — into simulation environments, data annotation quality assurance, and fine-tuning services — if they want to remain relevant as global competition intensifies and prices keep falling.
Where This Fits in India’s Broader Robotics Story
This human-data trend is unfolding against the backdrop of a robotics market that’s growing fast by any measure — estimates place India’s robotics sector growth between 17% and 34% year-on-year in 2026, making it one of the fastest-growing robotics markets globally, even from a relatively low base. Robot density in Indian factories remains far below markets like South Korea or Japan, but that gap is exactly why analysts see so much headroom for growth — in manufacturing automation, in agri-robotics, and now, in this less visible layer of data infrastructure.
India hasn’t yet produced a Tier-1 humanoid robotics company to rival the US or China. But this data-training wave suggests the country may be carving out a different kind of role in the global robotics value chain — not building the robot bodies, but supplying the human behavioral data that makes those robots intelligent. It’s a pattern with an obvious parallel: much like India’s IT services boom of the 2000s turned the country into the world’s back office for software, its labor pool may be positioning it to become the world’s back office for physical AI training data.
What to Watch Next
For anyone tracking India’s robotics and automation sector, the data-training economy is worth watching closely over the next 12-18 months. Key questions include whether Indian firms can move beyond commoditized video capture into higher-value simulation and annotation work, whether domestic robotics companies start tapping this same talent pool for their own humanoid development, and whether falling per-contract prices push more consolidation among data-collection startups.