Charting the Future of Earth Observation 2024

Charting the Future of Earth Observation: Technology Innovation for Climate Intelligence10Technology pipeline: In-orbit data processing with satellite edge computing In-orbit data processing represents a significant advancement in the management and use of EO data. Satellite edge computing processes EO data directly onboard the satellite, integrating advanced computational capabilities. Edge computing can also facilitate spike-based processing, whereby the satellite processes and downlinks data only when its sensors detect a change in the surrounding environment. This helps in extracting higher performance on low power and storage, increasing the per orbit efficiency by smart capture, processing and dissemination. By shifting data processing from ground stations to the satellites themselves, this approach greatly reduces latency and the need for data download. SkyServe, an Indian space technology start- up focused on onboard AI and edge computing for satellites, recently demonstrated the delivery of real-time insights using optical imagery on their Mission Matterhorn. The SkyServe STORM system was successfully deployed in orbit aboard D-Orbit’s ION Satellite Carrier, a fleet of orbital transfer vehicles. Using D-Orbit’s live EO data feed, on- board computer and data distribution resources, the STORM satellite edge computing system showcased its ability to process live images using DL models for real-time vegetation classification. This demonstration is crucial for disaster response as it significantly enhances the efficiency and speed of data processing, allowing faster transmission of vital information to emergency responders. SkyServe is further scaling the solution for multi- sensor fusion of multispectral and hyperspectral data from space by partnering with Loft Orbital. SkyServe’s Mission Denali will install the SkyServe STORM platform on the YAM-6 spacecraft, Loft Orbital’s first spacecraft capable of performing “virtual missions”. By decoupling the concepts of “satellites” (a platform with payload hardware) and “missions” (an objective executed by the payload), the SkyServe STORM platform aims to scale the practical application of in-space data processing by executing geospatial AI models on a whole host of satellites. Demonstration of satellite edge computing with EO data FIGURE 4 Note: Live satellite EO imagery (left) from D-Orbit ION carrier and the output from SkyServe STORM platform showing near real-time vegetation classification (right). Source: SkyServe. (2024). STORM onboard D-Orbit’s ION SCV-004. https://www.skyserve.ai/missions/matterhorn.