Thanks to NVIDIA Jetson, Planet achieves AI object detection on its Pelican satellite
(Rather than sending heavy images then using them on the ground, this allows already processed and much lighter images to be sent to the ground, saving time)Â
SAN FRANCISCO, April 7, 2026 – Following the announcement of a strategic initiative with NVIDIA, Planet Labs PBC (NYSE: PL) announced a major technical advancement: the successful deployment and execution of direct AI-driven object detection aboard its Pelican-4 satellite, paving the way for on-orbit computing for rapid analyses, a capability Planet calls “Planetary Intelligence.”
On March 25, 500 km above Alice Springs, Australia, Planet’s Pelican-4 captured an image of an airport, then successfully used its onboard NVIDIA Jetson Orin module to run an AI model to detect aircraft in moments. This represents one of the first times an Earth observation satellite has gone beyond simple data capture to perform AI inference and analysis directly on board.
“This success provides a glimpse into the future of what we call Large-Scale Planetary Intelligence” said Kiruthika Devaraj, vice president of Avionics and Spacecraft Technologies. “By running edge AI on the NVIDIA Jetson platform, we can reduce the time between when we “observe†a change on Earth and when a client can “take action†, while minimizing latency and transmission costs to the ground. This move to edge-integrated AI is a technological advancement that can differentiate solutions like Planet’s Global Monitoring Service (GMS), providing valuable insights to our customers and enabling rapid response times when it matters most.»
“This step with NVIDIA can accelerate the pace of obtaining information, reducing the time needed to obtain potential answers from hours to minutes.” It can make a critical difference to our customers, from disaster response to security and beyond », a déclaré Will Marshall, PDG et cofondateur de Planet. « More broadly, this is an exciting step towards implementing Planetary Intelligence. We are taking AI from the Internet to the physical world, effectively connecting the “eyes†of our satellites to an onboard “brain†to create a nervous system for the planet. »
This advancement may enable Planet’s Pelican constellations as well as future Owl constellations to evolve into a near real-time intelligence network. By leveraging NVIDIA Jetson and high-speed inter-satellite links, Planet works to effectively reduce latency. Rather than waiting for data to be downloaded and then analyzed on the ground, customers could receive actionable insights within minutes of capture. The end-to-end process, covering initial data generation, object detection by deep neural networks and full georectification, is designed to be carried out entirely in orbit. By producing analyzes in GeoTIFF (georeferenced image – pixels) and GeoJSON (geographic shapes – objects) format within spatially isolated Docker containers, Planet aims to be able to generate information about events in minutes.
Although onboard models are still in their early stages and will continue to be improved, the successful integration of AI applications in space paves the way for a new era of flexibility for satellite constellations. By placing this computing power directly on the satellite, Planet aims to unlock the future of Earth intelligence on a planetary scale with near real-time analytics.
À propos de Planet Labs PBC
Planet is a leading provider of daily global satellite imagery and geospatial solutions. Planet is driven by the mission to image the Earth every day and make changes visible, accessible and actionable. Founded in 2010 by three NASA scientists, Planet designs, builds and operates the largest fleet of Earth observation satellites. Planet provides critical data, advanced analytics and software solutions to customers including the world’s leading companies in agriculture, forestry, intelligence, education and finance, as well as government agencies, enabling users to simply and efficiently derive unique value from satellite imagery. Planet is a mission-driven company listed on the New York Stock Exchange under the symbol PL.
Source: NVIDIA
