Mecka AI, a New York City-based startup that trains robots using human data collected from body sensors and iPhones, has raised $60 million in funding. The round includes a $25 million Series A, signaling strong investor appetite for novel approaches to robotic training.
The company taps into a vast reservoir of human motion—hand gestures, walking gaits, and other physical behaviors—captured via wearable sensors and smartphone cameras. This data serves as a foundation for teaching robots to replicate nuanced human movements, a task that traditional programming has struggled to achieve.
The $60 million figure encompasses both the Series A and additional capital, though the exact breakdown beyond the $25 million Series A was not disclosed. The funding will likely accelerate Mecka AI's data collection efforts and expand its engineering team as it competes in the crowded robotics AI space.
By ingesting real human motion data, Mecka AI aims to overcome one of robotics' enduring challenges: teaching machines to move with the fluidity and adaptability of people. If successful, its technology could advance applications from industrial automation to assistive devices.
Critics argue that reliance on human-sourced data may limit scalability, as collecting diverse motion samples across populations remains expensive and time-consuming. The approach also raises privacy questions around continuous body-sensor monitoring.