NASA tests AI medic for deep-space missions

NASA's Johnson Space Center is testing the Crew Medical Officer Digital Assistant (CMO-DA), a clinical decision support system intended to help crews diagnose and treat medical symptoms during long-duration missions, The Register reported June 27. Per a Red Hat blog post (June 24, 2026), the CMO-DA has migrated from a cloud-dependent proof-of-concept to a fully disconnected, edge deployment running on a terrestrial twin of HPE's Spaceborne Computer. Red Hat stated the deployment uses RamaLama, an open source project the company supports, to run and serve AI models in isolated, containerized environments.

Per Red Hat's blog, RamaLama packages models as Open Container Initiative-compatible images so they run predictably on diverse hardware, including edge servers. The CMO-DA performs multimodal inference, combining large language models for medical reasoning and vision-language models for image-based symptom analysis. A Google Cloud blog post from August 2025 -- predating the RamaLama integration -- adds that initial trials evaluated CMO-DA outputs using the Objective Structured Clinical Examination framework, and that Google and NASA collaborated with medical doctors to refine the model. Red Hat's June 2026 post indicates the next iteration will integrate Red Hat Enterprise Linux AI (RHEL AI) for a hardened, stable runtime stack.

Communication latency for Moon and Mars missions, and possible blackout windows, is a core driver for autonomous clinical decision support. Containerized model delivery and on-device inference are an established pattern for deployments where intermittent connectivity and limited power make continuous cloud access infeasible. The use of OCI-style containers for model portability and an HPE Spaceborne-style edge appliance mirrors practice in other sectors requiring auditable, reproducible AI at the edge.

Red Hat said the system will be demonstrated to NASA leadership after Earth-based validation. Three signals matter: whether CMO-DA moves from terrestrial twin testing to an on-orbit trial; whether the project transitions to RHEL AI as Red Hat indicated; and whether the teams publish validation metrics and audit trails reviewable by clinicians and regulators. For practitioners outside aerospace, the project is a case study in shipping auditable, offline clinical decision support to remote or resource-constrained settings.

The Register (June 27, 2026) is the primary independent reporting outlet. Technical and program details also derive from Red Hat's blog and Google Cloud's August 2025 post; neither provided peer-reviewed clinical outcomes in the public posts, and NASA has not released a detailed public clinical validation report in the sources reviewed.