Planet Labs has been nominated for the 2026 World AI Awards in the Aerospace category, recognising the Earth-imaging company’s work combining satellites, artificial intelligence and machine learning to turn enormous volumes of imagery into information that organisations can actually use.
Planet’s approach to AI is unusually physical. Instead of keeping every algorithm inside a terrestrial data centre, the company is beginning to put AI computing directly aboard its satellites. The goal is straightforward: analyse useful information closer to where the imagery is captured and reduce the delay before it reaches people on Earth.
The World AI Awards recognises organisations, individuals, products and technologies contributing to the development and real-world application of artificial intelligence across industries.
Planet operates Earth-observation constellations that repeatedly photograph the planet. That creates a valuable archive, but also a problem. Humans cannot realistically inspect every image looking for a newly built road, damaged infrastructure, a disappearing section of forest or another change on the ground.
AI helps narrow that gap.
Planet puts AI processing directly into orbit
Planet’s next-generation Pelican satellites provide perhaps the clearest example.
The Pelican constellation delivers high-resolution Earth imagery, with the satellites carrying NVIDIA Jetson AI platforms for onboard edge computing. Instead of automatically sending every piece of raw information back to Earth before analysis begins, onboard computing can process imagery in orbit and identify useful information earlier.
That idea moved beyond theory in 2026.
Planet reported that its Pelican-4 satellite successfully ran automatic object detection in orbit using imagery captured over Alice Springs, Australia, on March 25. According to the company, the AI processing occurred only minutes after the imagery was collected.
Planet described it as the first successful execution of AI-driven object detection directly aboard one of its spacecraft.
The practical issue here is latency. Satellite imagery can be extremely valuable during fast-moving situations, but less so if analysis arrives too late. Processing selected information in orbit could shorten the route between observing something and identifying what matters.
Planet is exploring applications including object detection, vegetation and crop classification, security monitoring and disaster response. These capabilities should still be distinguished from a universal real-time AI service: onboard processing is developing alongside the expanding Pelican constellation.
Machine learning searches for changes humans could miss
Not all of Planet’s AI work happens in space.
Its Road and Building Change Detection products use deep learning to identify development across large geographic areas. Planet’s technical documentation says the system applies a modified U-Net semantic segmentation model to PlanetScope imagery, classifying pixels as roads, buildings or neither.
The results are aggregated over time, allowing another predictive model to identify significant differences between expected and observed conditions.
That means an analyst does not have to manually search through an enormous stream of satellite images just to discover where construction occurred.
Planet says the system can provide weekly information showing where roads and buildings are appearing across country-scale areas. Potential uses include infrastructure monitoring, development tracking and identifying activity in protected areas.
It is a good example of AI doing the repetitive work first, then pointing human analysts toward the places that deserve attention.
AI turns satellite imagery into forest intelligence
Forests create another difficult monitoring problem. Measuring canopy height, forest loss and stored carbon across enormous areas using field surveys alone is expensive and slow.
Planet’s Forest Carbon products use machine learning and multiple Earth-observation datasets to estimate aboveground carbon, canopy height and canopy cover.
The company says its deep-learning models are trained and validated using airborne LiDAR alongside satellite observations. Planet also incorporates measurements from NASA’s Global Ecosystem Dynamics Investigation, or GEDI, with additional satellite datasets to build broader estimates of forest conditions.
The result is designed for applications including deforestation monitoring, carbon-project measurement and reporting, reforestation analysis and supply-chain risk assessment.
Planet’s Forest Carbon Monitoring product provides quarterly observations at 3-metre resolution, while its wider datasets allow changes to be examined from regional scales down toward individual forest disturbances.
Again, AI is not the end product. The useful output is information about what is physically changing on Earth.
Graham Cooke, President of the World AI Awards, said:
“Planet Labs’ nomination highlights an exciting direction for artificial intelligence in aerospace: moving intelligence closer to where data is actually being created.
“From running object-detection models aboard satellites to using machine learning to identify infrastructure and forest change across huge areas, Planet is showing how AI can help turn Earth-observation imagery into information people can act on more quickly. We congratulate the Planet Labs team on its 2026 World AI Awards nomination and look forward to following the development of its AI-enabled satellite capabilities.”
Planet Labs joins organisations, researchers, entrepreneurs and technology developers being recognised through the 2026 World AI Awards.
The programme recognises organisations, individuals and technologies contributing to the development and application of artificial intelligence across industries, including sectors where AI is increasingly connected to sensors, machines and physical infrastructure.
Planet Labs offers a particularly tangible version of that shift. Its satellites generate the observations. Machine-learning models help find meaningful patterns inside them. Now, with the Pelican programme, part of that computation is beginning to move into orbit itself.
There is still a distinction between what is operational and what Planet intends to build next. Its machine-learning analytics and Earth-observation products are already available, while broader onboard AI capabilities are developing as the Pelican constellation expands.
That progression could matter well beyond satellite imaging. When spacecraft can analyse selected observations before transmitting them, the satellite becomes more than a camera in orbit. It starts becoming part of the intelligence pipeline.
Learn more about Planet Labs and its Earth-observation technology at planet.com.
Discover the World AI Awards 2026, explore the nominees and learn more about the awards at worldawards.ai.
