NASA's Dexterous Robotics Team at Johnson Space Center is building the robotic systems designed to work in environments built for humans — turning latches, moving cargo, inspecting equipment. The 16-member group, led by Shaun Azimi, sits within the Robotic System Technology Branch and draws its lineage from two flagship humanoid platforms: Robonaut 2, which spent seven years aboard the International Space Station, and Valkyrie, NASA's first bipedal humanoid robot. Many current team members cut their teeth on those programs. The team's core thesis is augmentation, not replacement. "Our team is not trying to replace human explorers with robots but instead make human exploration safer and more sustainable," Azimi said. The focus is dexterous manipulation — tasks humans do with their hands — in extreme environments where crew time is scarce and risk is high. The operational logic is straightforward: send more capable robots ahead or alongside, reduce EVA risk, extend what a small crew can accomplish. The most tangible output is the Integrated Mobile Evaluation Testbed for Robotics Operations (iMETRO) facility at Johnson. iMETRO combines open-source software, simulation assets, space vehicle and habitat mockups, a selection of "house robots," and an outdoor rock yard. It is available to NASA programs and external partners alike. The facility's value proposition is specificity: it shows technology providers exactly what tasks NASA needs done, eliminating guesswork and enabling co-design between roboticists and habitat architects. Real demonstrations have already run through iMETRO. A team from PickNik Inc. tested software enabling a robotic arm to recognize a spacecraft hatch, turn the latch, grasp the handle, open the door, and transfer cargo bags. A NASA intern developed and tested software using a commercial robotic arm and camera to inspect and maintain a cold stowage freezer like those aboard the station. These are unglamorous logistics tasks — exactly the kind robots need to handle reliably before anyone trusts them with more. The team is organized into mechatronics and software subgroups, but most members have cross-functional experience spanning electronics, mechanics, and simulation software. This generalist-specialist hybrid structure reflects the reality of space robotics: the hardware-software boundary is artificial when your robot must operate autonomously in a habitat designed for human hands. The team also collaborates with private industry, including an oil and gas company seeking robotic technologies for harsh environments. Near-term work targets sustained human presence on the lunar surface, with explicit connections to Moon Base planning. But Azimi noted the technologies have Mars applications, and the team is collaborating on a forthcoming NASA challenge inviting public ideas for Mars exploration technology. The through-line is clear: prove dexterous manipulation in lunar habitats, then extend the capability envelope outward. Azimi frames Johnson's robotics niche as the human-environment intersection — robots that work in spaces designed for people, alongside people. "We're uniquely positioned to bring in folks who are designing the human environments," he said. The bet is that co-designing robots and habitats simultaneously, rather than bolting robots into finished spaces, produces better outcomes for both.