VC Eclipse has a new $1.3B to back — and build — ‘physical AI’ startups

Eclipse Ventures, a Palo Alto-based venture capital firm, is making waves in the investment landscape with its recent $1.3 billion funding aimed at fostering the burgeoning field of 'physical AI.' This significant capital infusion includes a $591 million fund dedicated to early-stage startups, alongside a separate fund focused on growth-stage companies. The firm has strategically invested in a diverse array of companies that are pushing the boundaries of technology in the physical realm. Notable investments include Arc, an electric boat manufacturer; Redwood Materials, a company specializing in battery recycling; Bedrock Robotics, which is innovating in self-driving construction vehicles; Wayve, a player in autonomous vehicle technology; and Mind Robotics, an industrial robotics lab. Jiten Behl, a partner at Eclipse, emphasized the transformative potential of this new era in technology, stating, "This is the first time where stuff is going to move from our screens into the physical world; we’re going to see advanced levels of intelligence, along with actual actions, in terms of solving problems in the real, physical world." This shift is being fueled by a unique blend of talent, technological progress, market demand, and supportive policies, all underpinned by substantial financial resources. Eclipse's investment strategy is noteworthy as it aims to create a cohesive ecosystem of startups across various sectors, including transportation, energy, infrastructure, computing, and defense. Behl described this approach as essential for achieving scale and fostering collaborations among portfolio companies, enabling them to leverage each other's strengths. Moreover, Eclipse is not just an investor but also a builder, with plans to incubate new companies within its framework. Behl hinted at exciting developments already in the works, focusing on ideas that transcend individual sectors. He posed critical questions about connecting these sectors and utilizing data to enhance AI models, which he believes will serve a wider audience. "That’s sort of the general thesis that we’ve been working on," he concluded, signaling a promising future for 'physical AI.'