Mecka AI, a New York City-based startup, has raised $60 million through two undisclosed fundraises to advance its robotics training using human motion data. The company closed a $25 million Series A round in November followed by a $35 million follow-on investment, with Framework Ventures leading the rounds. Other investors include Menlo Ventures, SV Angel, Kindred Ventures, and angel investor Ted Xiao, according to fortune.com.
The startup focuses on collecting embodied AI data such as hand gestures and walking gaits from body sensors and iPhones to improve robot training. CEO Josh Gao highlighted the growing importance of robotics in the AI landscape during a video call from Shenzhen, where he was visiting a factory. Gao declined to disclose the valuation at which the capital was raised. The funding aims to leverage vast reservoirs of physical motion data to enhance robotic capabilities.
Embodied AI data is increasingly recognized as crucial for robotics development, with companies like Wayve using visual inputs for autonomous driving and MicroAGI recording home cleaning activities for training. Mecka AI’s approach to sourcing human motion data from wearable sensors and smartphones positions it within a competitive but expanding market focused on improving robot learning through real-world human actions. The involvement of investors with backgrounds in crypto and AI underscores the sector’s growing appeal.
Mecka AI’s funding rounds, led by Framework Ventures, mark a significant step in robotics data training, with the Series A round completed in November and the follow-on investment closing recently. The startup’s next milestones will likely involve scaling data collection and refining robot training models using the amassed human motion datasets.