# NASA wants robot helpers in space Canonical URL: https://www.theroboteconomy.news/articles/nasa-google-bag-moving-baseline/ Publisher: The Robot Economy Editorial byline: Imani Reed, Frontier Desk Published: 2026-10-29T02:06:36.835Z Updated: 2026-10-29T08:44:32.439753Z Editorial information: https://www.theroboteconomy.news/about/#editorial-desks An 85% laboratory baseline is a measurable start. Missing trial counts and recovery data limit what it says about dependable spacecraft assistance. ## Summary Google reports an 85% baseline for moving a medical bag at NASA’s iMETRO facility. The new task result is distinct from mission qualification: trial count, recovery burden and local-device performance remain unreported in the announcement. ## Key facts - Google published the proof-of-concept account on 27 October 2026. - The reported 85% baseline concerns locating and moving a medical bag in early laboratory conditions. - Trial count, recovery labour and mission qualification are not established by that percentage. ## Lead image Image: https://www.theroboteconomy.news/images/articles/nasa-google-bag-moving-baseline-imetro.webp Description: A robotic arm holds a cargo bag beside spacecraft hatch mockups in NASA’s iMETRO laboratory. Context photograph of the facility, not an identified frame of Google’s 85% test. Credit: Context photograph: NASA. Cargo-handling hardware at iMETRO, published September 2026; not an identified image of Google’s trial. ## Report Google Public Sector has reported an initial 85% success rate for a robot locating and moving a medical bag in early laboratory tests at NASA’s Johnson Space Center. Its October 27 account describes a proof of concept using the Pi0 vision-language-action model, which connects instructions and visual input to robot movements. The result adds a task-level number to a programme previously described largely through its facilities and demonstrations. Sources: [s1]() [s3]() The useful boundary is the task being counted. Moving a bag is a logistics experiment, not evidence of autonomous diagnosis, ultrasound operation or a robot qualified for a lunar mission. Google describes retrieving and preparing an ultrasound machine as a possible assistance scenario; broader voice-directed tasks remain an ambition. Human medical judgement stays central to its proposed system. Sources: [s1]() The account combines a camera positioned outside the robot’s hand with simulation in MuJoCo, a physics engine, before hardware work. It also reports fine-tuning a four-billion-parameter model in 2.5 hours on Google Cloud G4 computing instances. That is a training result, not a disclosed local inference speed, energy budget or end-to-end task time. Sources: [s1]() Google says fine-tuned models can operate locally and that the team evaluated about 15 edge-computing devices over two years. This matters because spacecraft cannot depend on continuous Earth data-centre access. But the announcement does not identify a selected flight device or provide matched power, latency and task-success measurements for it. Sources: [s1]() NASA’s separate account of its Integrated Mobile Evaluation Testbed for Robotics Operations, or iMETRO, explains why the setting matters. Physical mockups, simulation and robots let developers test logistics and maintenance around human-designed spacecraft interfaces. That context supports a laboratory integration story; it does not independently replicate Google’s percentage. Nextgov/FCW also obtained a NASA statement describing ongoing assessment of potential applications. Sources: [s2]() [s3]() The next useful evidence is a trial ledger: attempts, successful completions, interventions and recovery time, with the camera, hardware and local-computing configuration specified. Without the trial count, 85% alone cannot establish statistical precision or dependable crew-time savings. The advance is a reported manipulation baseline; operational reliability remains the next test. Sources: [s1]() [s2]() ## Conclusion Separates the newly reported task baseline, cloud training duration and proposed local execution into three different acceptance questions. Identifies missing trial/recovery and matched device measurements without converting demonstrations into mission readiness. ## Corrections 2026-10-29T08:21:36.289738Z: Replaced the reused Valkyrie photograph with a different NASA photograph of the iMETRO facility and sharpened the headline. The reporting and original publication date are unchanged. 2026-10-29T08:44:32.439753Z: Shortened the headline at the owner’s request. The reporting, photograph and original publication date are unchanged. ## Disclosures Sponsored: no Desk bylines, AI assistance and editorial policies: https://www.theroboteconomy.news/about/ ## Sources - s1: [Google Cloud and NASA: Testing robotic systems for deep space missions]() — Google Cloud; company-announcement; 2026-10-27T00:00:00Z - s2: [NASA, Google test AI-powered medical robot for deep space missions]() — Nextgov/FCW; independent-reporting; 2026-10-28T16:16:00Z - s3: [Reliable Robots: Meet Johnson’s Dexterous Robotics Team]() — NASA; official-record; 2026-09-28T00:00:00Z - s4: [NASA Images and Media Usage Guidelines]() — NASA; official-record