Firefly Aerospace Operates NVIDIA Jetson Platform in Lunar Orbit for First Time

Firefly Aerospace is deploying edge computing hardware from NVIDIA within its Ocula space imaging service to perform on-orbit data processing. This integration targets the aerospace and orbital infrastructure sectors by enabling autonomous data analysis in cislunar space.

Cooperation Rationale and Challenges
Space-based sensing operations traditionally rely on transmitting large volumes of raw data over constrained radio frequency bandwidths. During a previous lunar mission, lander systems downlinked 120 gigabytes of raw sensor data, requiring extended terrestrial CPU processing. The integration of edge AI systems addresses the latency and bandwidth limitations inherent in deep space communications by eliminating the requirement to transmit uncompressed raw files to Earth.



A rendering of the Blue Ghost Mission 2 lander on the lunar surface. Image courtesy of Firefly.

Technical Solution and System Architecture
The technical architecture combines the NVIDIA Jetson edge computing module with artificial intelligence software developed by SciTec, a subsidiary of Firefly Aerospace. This system is integrated into the Ocula sensor suite, which captures imagery across visible and ultraviolet spectrum bands. Powered by solar arrays, the hardware runs AI inference algorithms directly on the spacecraft. The system processes the raw optical data locally, extracts predefined analytical parameters, and transmits only the compressed, relevant data packets to ground stations.



Blue Ghost Mission 2 structure testing at NASA’s Jet Propulsion Laboratory. Image courtesy of NASA/JPL-Caltech.

Deployment and Implementation
The system is scheduled for deployment during the Blue Ghost Mission 2. The computing hardware and Ocula sensor will operate aboard the Elytra spacecraft, which will maintain a lunar orbit for a five-year operational period. While a separate lander descends to the lunar surface with independent scientific instruments, the Elytra vehicle will function as an orbital processing node, executing the AI-powered data processing chain continuously.



The Rashid Rover 2 that will fly onboard Blue Ghost Mission 2. Image courtesy of Firefly.

Applications and Industrial Use Cases
The on-orbit processing capability serves the aerospace, defense, and orbital resource extraction industries. Specific operational use cases include high-resolution mapping of lunar landing zones for automated navigation systems and the detection of mineral deposits, such as ilmenite, for future energy applications. The system also provides situational awareness by monitoring orbital infrastructure, tracking cislunar objects, and assessing surface operations for institutional clients.

Expected Impact and Results
By executing data analysis at the edge, the system reduces downlink latency from weeks to near real-time transmission. This architecture decreases bandwidth consumption and lowers the operational costs associated with deep space network utilization.

Edited by Natania Lyngdoh, Induportals editor, assisted by AI.

www.nvidia.com