Bio:
Chao Cao is the co-founder and CEO of Sancho Robotics, which builds general-purpose robots for advanced manufacturing and biopharma. He holds a Ph.D. in Robotics from Carnegie Mellon University, where he led the planning team in the DARPA Subterranean Challenge, earning 1st- and 2nd-place finishes and the Most Sectors Explored award. His research has been published in Science Robotics and received the IROS 2022 Best Student Paper Award, as well as the RSS 2021 Best Paper and Best Systems Paper Awards, the first time a single paper won both in the conference's history. Before Sancho, he was a research scientist at the Boston Dynamics AI Institute working on whole-body dexterous mobile manipulation.
Abstract:
Today’s advanced factories and labs are packed with specialized automation machines, yet skilled workers are still needed to tend, operate, and connect them. Every machine is automated except the gaps between them. These gaps are the last mile of automation: the missing piece that stops us from running fully automated workflows, and one of the hardest open problems in robotics. Closing them takes robots that are precise, reliable, and adaptable to new tasks in unstructured, contact-rich environments.
In this talk, I’ll share how we’re building a “physical API” at Sancho Robotics: general-purpose robots that connect islands of automation into end-to-end workflows, including work featured in the NVIDIA GTC 2026 keynote. I’ll also share my path from the DARPA Subterranean Challenge at CMU to the Boston Dynamics AI Institute to founding a startup, and what changes when your robot has to work every day, not just once for the camera.