Humyn Labs Raises $47 Mn To Collect Training Data For Robots

SUMMARY

Physical AI startup Humyn Labs has raised $47 Mn in a funding round, comprising $26 Mn in equity led by Accel and $21 Mn in venture debt led by Lighthouse Canton.

Humyn claims to have crossed $50 Mn in annualised revenue within six months of starting operations.

The startup will use the capital to expand its robot deployment operations to Brazil, South Korea and India, strengthen its data collection hardware.

Physical AI startup Humyn Labs has raised $47 Mn (about ₹439 Cr) to expand its data collection and robot deployment operations. The round comprises $26 Mn in Series A equity led by Accel and $21 Mn in venture debt led by Lighthouse Canton, with participation from Trifecta Capital.

Other equity investors included Prosus, BeeGlobal, Bajaj Finserv Intelligence, Bajaj Alts AI Opportunities Fund, Dallas Venture Capital, Jump Trading LLC, and angel investor Anita Pramoda.

Founded earlier this year by former Nazara CEO Manish Agarwal and Ishank Gupta, Humyn Labs collects data across more than 20 countries. It works with AI research labs, robotics companies, and developers of world models, which simulate how physical environments change in response to actions. The startup claims it crossed $50 Mn in annualised revenue within six months of starting operations.

The San Francisco-based startup plans to expand its robot deployment operations to Brazil, South Korea, and India. It also plans to strengthen its data collection hardware and processing capabilities and build out its presence in San Francisco, London, and Seoul.

Humyn Labs claims to have crossed $50 Mn in annualised revenue within six months of starting operations.

The startup plans to expand its robot deployment operations to Brazil, South Korea and India. It will also strengthen its data collection hardware and processing capabilities and expand its presence across San Francisco, London and Seoul.

App Launched

Physical AI refers to AI systems that interact with the real world through robots and other machines. Unlike conventional AI models that primarily process digital information, these systems need to understand how objects move, how humans perform tasks and how to respond to changing physical environments.

For instance, a warehouse robot must learn to pick up different objects, navigate obstacles and handle materials without requiring instructions for every movement.

Humyn Labs aims to address this challenge by recording how people perform tasks across factories, warehouses, farms and homes. Its proprietary devices capture human movements and interactions, which are then converted into labelled datasets to train robotics models.

The startup also works with robotics developers to test robots in real-world environments and improve their performance.

India’s Physical AI Opportunity And The Questions Around It

The opportunity extends beyond building robots. According to a July 2026 report by Counterpoint Research, India’s data collection and annotation services for physical AI could unlock a $10 Bn revenue opportunity by 2030.

The country’s large workforce, diverse working environments and relatively low data collection costs could make it an attractive destination for companies seeking real-world training data for robotics systems.

However, the business also raises questions around privacy, consent and the ownership of training data.

Physical AI datasets can involve recordings of people inside their homes, factories and workplaces. Such footage may capture personal information, household activities, workplace processes and commercially sensitive details.

This makes informed consent, data storage and restrictions on how recordings are shared particularly important. Questions also remain around how workers are compensated and whether they retain any control over footage used to train commercial AI systems.

There is a broader labour question, too. The workers whose movements are being recorded are helping train systems that could eventually automate some of their jobs. Warehousing, packaging, assembly-line operations, cleaning and inventory management are among the activities that physical AI-powered robots could increasingly take over.

This creates a potential contradiction at the heart of the emerging industry: human labour is becoming an important input in developing technologies designed to reduce dependence on human labour.

Leave a Comment