AI Processing of Earth Images Can Now Run in Space

2026-05-01 · Source: IEEE Spectrum · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems, Emerging Technologies & Innovation · Depth: Intermediate, medium

Summary

Planet Labs has successfully achieved AI image processing aboard its Pelican-4 multispectral satellite, marking a significant milestone for the Earth observation industry. The satellite, deployed in 2025, identified and highlighted over a dozen aircraft at an airport in Alice Springs, Australia, using an onboard AI model. This capability, developed over 18 months by Planet Labs engineers, aims to overcome data latency challenges where traditional ground processing can take 6-12 hours. The company's constellation of Dove, SuperDove, and Pelican satellites generates 30 terabytes of data daily. The onboard Nvidia Jetson Orin GPU modules on Pelican satellites can analyze a 16,000-pixel image in half a second, delivering insights to users in minutes. Planet Labs plans to expand this AI capability to its new Owl satellites and envisions "planetary intelligence" with future large language models (LLMs) running in space.

Key takeaway

For Computer Vision Engineers developing Earth observation systems, Planet Labs' success with onboard AI processing on Pelican-4 satellites demonstrates a viable path to overcome data latency. You should explore integrating edge AI capabilities directly into satellite payloads using robust, space-qualified processors like Nvidia Jetson Orin to enable near real-time object detection and classification, significantly enhancing the timeliness and utility of your remote sensing data.

Key insights

Onboard AI processing on satellites significantly reduces data latency for Earth observation, enabling real-time insights.

Principles

• Proximity of compute to sensors minimizes data transfer delays.
• Reliable AI in space requires robust algorithms for raw data.

Method

AI image recognition algorithms run on onboard Nvidia Jetson Orin GPUs to process raw satellite imagery, classifying objects and delivering results within minutes of image capture.

In practice

• Deploy Nvidia Jetson Orin GPUs for space-based AI.
• Develop algorithms robust to satellite motion and vibration.
• Consider LLMs for future text-based satellite insights.

Topics

• AI in Space
• Earth Observation
• Planet Labs
• Pelican Satellites
• NVIDIA Jetson Orin

Best for: Computer Vision Engineer, AI Engineer, Machine Learning Engineer, AI Scientist

Related on AIssential

• Science Earth: Towards A Planet-Scale Operating System for AI-Native Scientific Discovery — A planet-scale runtime enables AI capabilities to self-organize for distributed, self-correcting scientific discovery.
• Making Sense of the Early Universe — AI and GPU acceleration dramatically enhance astronomical discovery by automating complex data analysis.
• How NVIDIA's Earth-2 uses AI to Accurately Predict Weather — NVIDIA Earth-2 provides an open, GPU-accelerated AI stack for faster, more accurate weather and climate forecasting.
• Using AI to scan the Earth — AI-powered dynamic targeting enables satellites to autonomously capture cloud-free Earth observation imagery.
• Robots’ Holiday Wishes Come True: NVIDIA Jetson Platform Offers High-Performance Edge AI at Festive Prices — NVIDIA Jetson developer kits offer scalable edge AI compute for diverse robotics and physical AI applications.
• AI galaxy hunters are adding to the global GPU crunch — New space telescopes are generating unprecedented data volumes, necessitating GPU-accelerated deep learning for astrophysical analysis.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by IEEE Spectrum.