Eka Robotics has raised a $150m Series A led by Andreessen Horowitz and F-Prime, the company said on 20 October. TechCrunch reports that the round values it at $1.45bn. Eka was co-founded by MIT professor Pulkit Agrawal and former Google DeepMind researcher Tuomas Haarnoja.
Many robot foundation models combine vision and language. Eka’s “vision-force-action” model adds a sense of touch: it reads force sensors in a custom gripper alongside cameras, and is trained largely with reinforcement learning in simulation, where a robot can gather thousands of hours of practice in minutes. Eka says its robots pick a wide range of objects, including ones they were not trained on, at about 1,500 an hour, roughly 2.5 times a human baseline of about 600. Those are the company’s own figures. It has also shown robots screwing in light bulbs and inserting wires.
Why it matters: handling work that needs a delicate touch is one of the main things still keeping robots out of many factory and warehouse jobs, and force sensing is one way to tackle it. Eka names no customers; TechCrunch reports undisclosed industrial pilots. Kris Hauser, a University of Illinois professor, told TechCrunch that open questions remain about reliability over a year on a real production line.
Read Eka’s announcement: Eka Robotics raises $150M (Eka Robotics)
