Instawork Instacore takes robot training beyond labs
Robots may be getting smarter, but there is still one big piece missing: enough real-world data. Instawork Instacore is designed to address that gap by putting a five-camera wearable system on workers and capturing how real tasks happen in kitchens, warehouses, hotels and manufacturing facilities.
Introduced in India by Instaworks Robotics Lab (IRL), the system is built around a simple idea. Instead of asking robots to learn only inside controlled environments, their training data can come from people doing real jobs in changing conditions.
Instawork Robotics Lab takes data collection beyond labs
Instacore combines five cameras with a power-efficient compute backpack designed for a full work shift. A stereo head camera captures depth, while a wide-angle chest camera records the wider working environment. Wrist cameras track hand position and movement, supported by multiple IMUs for SLAM and pose estimation.
That setup matters because physical tasks rarely happen in isolation. A person may change tools, move around a workspace or deal with a different environment from one location to another. Instawork says its Pros tested multiple versions of the system in homes and commercial workplaces to make sure it could be worn during an entire shift.
The system is also modular. It can be configured with UMI grippers or different camera setups based on partner needs.
Instacore wearable system connects workers and robotics
The Instacore wearable system also creates a link between robotics companies and Instawork’s network of more than 10 million skilled professionals. The company says workers and participating businesses can take part without upfront hardware costs or disruption to normal operations, while businesses are compensated for hours already being worked.
For Instawork India, the approach also ties robotics data collection to new income opportunities for workers and local businesses.
The real challenge is scale, not just compute
The numbers in the release point to the size of the data challenge. Robot training data grew from roughly 100,000 hours in 2024 to about one million hours in 2025. Projections put the figure at 20 million hours in 2026, but that is still described as less than 0.04% of what researchers believe may ultimately be needed.
That puts Instawork Instacore in a space where scale and variety are central. Its design aims to capture different tasks, tools and environments rather than relying on a narrow set of lab conditions.
Instawork India launch targets the physical AI gap
The company’s India strategy goes beyond hardware. Instawork says it is building AI and robotics data operations in the country, drawing on engineering and technology talent while creating opportunities around AI, robotics and the future of work.
The broader market figures cited in the release underline the opportunity. Goldman Sachs forecasts a $38 billion humanoid robot market by 2035, while Grand View Research projects the global data collection and labelling market at $17 billion by 2030. Deloitte estimates that nearly two million skilled physical AI jobs could remain unfilled by 2030.
Read More:
Kramer 1G and 10G AVoIP expands AV choices at InfoComm India 2026
Cyfuture CFO appointment brings Himanshu Gupta onboard
Seqrite uncovers the hidden threat behind Operation ShadowRecruit
2 Min read