In Ahmedabad, India, a factory is utilizing GoPro cameras worn by workers to collect egocentric data, which is vital for training robots. This innovative approach allows for the recording of tasks from a first-person perspective, providing the necessary data for robotics companies to enhance their AI systems. A report by Stellaris Venture Partners indicates that leading robotics labs will require between 100 million and 1 billion hours of egocentric data in the next 2-3 years. To meet this demand, Indian startups such as Humyn AI, FPV Labs, and Neo Cambrian are entering the market to build a robust data pipeline. Ishank Gupta, cofounder of Humyn AI, noted that training robots for specific tasks requires between 100,000 and 1 million hours of data, emphasizing the need for tailored data collection. Objectways, another player in this field, has shifted its focus to egocentric data, currently producing 1,000 hours of data per day to fulfill the rising demand. With the growing need for accurate and extensive data, these companies are expanding their operations across various regions, including India, Latin America, Europe, and Southeast Asia.