NASA's Jet Propulsion Laboratory (JPL) and Indian edge computing startup SkyServe are collaborating on a groundbreaking project. They are testing artificial intelligence models aboard a D-Orbit satellite. These models, created for NASA's New Observations Strategies Sensorweb, are designed for near real-time monitoring of various environmental phenomena. This includes crucial events such as wildfires, floods, and urban heat islands, supporting both scientific research and disaster response efforts.
In March, SkyServe successfully completed tests of JPL's AI models using its STORM edge-computing suite. This suite was installed on a D-Orbit ION Satellite Carrier in low-Earth orbit. "We deployed this software onboard the existing satellite and completed the mission," explained Vinay Simha, SkyServe's founder and CEO, to SpaceNews. "We uplinked it from the ground station and ran these AI models onboard. It’s a new way of Earth observation."
The primary focus of these models is preprocessing and optimizing water and vegetation observations for easier data download. A news release highlights the need for sophisticated edge computing technology to integrate and optimize AI models across diverse sensor configurations. Vishesh Vatsal, SkyServe's chief technology officer, stated, "SkyServe’s technology plays a pivotal role in streamlining AI model deployment across diverse satellite platforms, ensuring consistency and efficiency. Our platform bridges the gap between groundbreaking AI models and the hardware diversity of satellite systems, enabling a unified approach to Earth observation."
D-Orbit plans to equip its ION satellite carriers with SkyServe STORM in 2024. Furthermore, Loft Orbital has already installed the SkyServe edge computer on its spacecraft, showcasing the technology's expanding adoption. Another SkyServe STORM unit is currently operating as a hosted payload on a SatRevolution cubesat, launched in December via India’s Polar Satellite Launch Vehicle. Beyond STORM, SkyServe also provides Surge, a platform for ground-based development and testing of AI models.