Amid a cooling embodied intelligence sector, RoboScience announced on November 2 that it has closed a Series A2 round in the hundreds of millions of yuan. The round was jointly led by Amber Capital, Hongruida Investment, and Shanghai Semiconductor Industry Investment, with participation from multiple institutions and listed company platforms. The company was co-founded by Tian Ye and Shao Lin, both alumni of the Stanford AI Lab. Tian Ye built extensive experience in large-scale technology deployment at Apple, while Shao Lin conducted in-depth research on robot learning at the National University of Singapore—together forming a complementary blend of academic depth and engineering expertise. Drawing on Stanford-affiliated academic resources, the team follows a full-stack technology roadmap integrating software and hardware. It has independently developed the Visics general-purpose embodied large model to tackle body generalization challenges, and independently developed the RoboMirage physical simulation engine to build a "simulation + video" dual data flywheel—cutting the cost of acquiring a single training data sample to 1/20 to 1/200 of traditional solutions and substantially breaking through data scale and cost bottlenecks. Within one year, the company has established a full-stack closed loop spanning data, models, and bodies, launched its first wheeled humanoid general-purpose robot, REX G1, and partnered with Tencent Cloud to create the world's first cloud-based EaaS embodied intelligence service—forming three industrialization paths: end-to-end solutions for terminal scenarios, cloud model services, and on-device controllers. Funding from this round will be used for large model iteration, POC validation and scaled deployment in core scenarios, and high-end talent reserves. The team remains committed to supporting vertical scenario deployment with underlying general-purpose technologies, progressively advancing toward the realization of general-purpose embodied intelligence.