Firefly Aerospace will pioneer on-orbit artificial intelligence processing with its upcoming Blue Ghost Mission 2, scheduled for launch in late 2026. Building on the success of its March 2025 lunar landing, which returned nearly 120 gigabytes of raw imagery for terrestrial analysis, Mission 2 will deploy the company’s Ocula moon imaging service equipped with the NVIDIA Jetson edge AI platform. This integration marks the first time a Jetson module will operate in lunar orbit, fundamentally altering the traditional space data pipeline. Historically, spaceborne sensors collect telemetry and downlink massive datasets to Earth, where CPU-based processing often requires weeks to yield actionable intelligence. Ocula will invert this workflow by executing AI inference directly aboard Firefly’s Elytra spacecraft during its planned five-year orbital tenure. Powered by solar arrays and driven by AI software from Firefly’s SciTec subsidiary, the Jetson module will autonomously analyze ultraviolet and visible spectrum imagery, compress the results, and transmit only mission-critical insights back to Earth in near real time. This on-orbit computing model dramatically reduces radio bandwidth latency and lowers downlink expenditures while accelerating scientific and operational decision cycles. Mission 2 will utilize a dual-component architecture to maximize scientific return. The primary lander will descend to the lunar far side to host a radio telescope array supporting a NASA-funded, UC Berkeley-led initiative aimed at detecting faint electromagnetic signatures from the cosmic Dark Ages. Simultaneously, the Elytra spacecraft will maintain a stable orbit to run the continuous Ocula processing chain. Firefly intends to iterate on the sensor hardware across subsequent launches, with plans to incorporate next-generation platforms like the NVIDIA Space-1 Vera Rubin Module as they reach flight readiness. Jason Kim, CEO of Firefly Aerospace, describes the architecture as the foundational layer for a broader cislunar data network. By positioning computational workloads in space, the company aims to create a connectivity layer analogous to terrestrial transatlantic cables, interlinking orbital assets to support scalable exploration. Ocula’s operational scope includes high-resolution mapping of future landing zones, identification of resource-rich minerals like ilmenite, and real-time situational awareness of surface infrastructure, vehicles, and cross-mission activity. These capabilities address growing demand from NASA, the U.S. Space Force, and private sector entities focused on lunar resource extraction and permanent settlement. With NASA projecting approximately thirty robotic lander missions over the next decade, Firefly’s deployment of edge AI in cislunar space positions it at the center of an emerging computational infrastructure. Shifting analytical workloads from ground stations to orbital platforms establishes a resilient, low-latency data ecosystem designed to sustain humanity’s expanding presence beyond Earth.