Eka Robotics, which is developing a "vision-force-action" architecture for its physical AI models, raised a $150M Series A at a $1.45B post-money valuation

Robots can now grasp and manipulate objects with unprecedented speed and gentleness, making them more versatile for complex tasks.

Eka Robotics has secured $150 million in Series A funding, achieving a post-money valuation of $1.45 billion. The company is developing a "vision-force-action" architecture for physical AI models, aiming to significantly improve robotic manipulation capabilities. Founded by MIT professor Pulkit Agrawal and former Google DeepMind researcher Tuomas Haarnoja, Eka's approach integrates force sensitivity with deep learning, allowing robots to adapt to diverse objects and tasks with remarkable speed and precision. Their system can reportedly pick up approximately 1,500 objects per hour, exceeding human performance, with a reliability rate above 95%. This advancement contrasts with traditional vision-language-action models by focusing on the interplay of force, vision, and action, and utilizes extensive physics simulation for reinforcement learning at scale. The funding round was co-led by venture funds a16z and F-Prime, signaling strong investor confidence in Eka's novel approach to physical AI.

Eka Robotics' "vision-force-action" model represents a significant departure from current generative AI robotics, which often prioritize planning over nuanced physical interaction. By incorporating force feedback directly into the learning loop, Eka addresses a critical gap in robot dexterity, moving beyond simple object recognition to fine-grained manipulation. This focus on force and touch, rather than solely vision, is a key differentiator that has impressed industry experts and may signal a new direction for physical AI development.

The $150 million Series A funding, led by prominent venture firms, highlights investor enthusiasm for solutions that bridge the gap between general AI capabilities and real-world physical tasks. Eka's success in achieving superhuman pick-and-place speeds with high reliability suggests a pathway to more robust and capable robots in industrial settings. While challenges remain in scaling these solutions to full-scale production lines and ensuring long-term durability, Eka's progress indicates a potential acceleration in the adoption of advanced robotics for a wider range of applications.