The mission to create a searchable database of Earth's surface | Will Marshall

Episode Summary

Executive Summary: Will Marshall describes Planet’s evolution from a bold mission to image Earth daily into a working global satellite system that now captures the planet every day. He argues that timely imagery plus AI can turn Earth into a searchable database, enabling practical action on agriculture, security, disaster response, deforestation, and environmental accountability.

Main Topics: Mission One: Daily Earth Imaging (Priority: 5/5): Planet’s original goal was to launch enough satellites to image the entire Earth every day and make the imagery broadly accessible, replacing outdated satellite photos with current views of change on the ground. Building the Satellite Constellation (Priority: 5/5): Marshall explains the scale and difficulty of deploying the fleet: many rocket launches, more than 200 satellites, and a global ground network to downlink and process imagery. Real-World Uses of Fresh Imagery (Priority: 4/5): He highlights how governments, businesses, and NGOs use Planet imagery for crop monitoring, mapping, disaster response, border security, refugee tracking, and deforestation detection. Planet Stories and Public Access (Priority: 4/5): Planet Stories opens the archive to anyone, letting users compare dates, create time-lapses, and share visual evidence of change for journalism, nonprofit work, or personal use. Mission Two: Space Plus AI (Priority: 5/5): The next phase is using AI to identify objects in imagery—ships, trees, roads, buildings, vehicles—so the data becomes searchable and structured rather than just visual. Queryable Earth (Priority: 5/5): Marshall’s long-term vision is an interface where users can ask questions about Earth itself—count houses, track trees, or detect deforestation over time—without manually inspecting images.

Key Arguments: Old satellite imagery is insufficient because human activity changes on timescales of days, weeks, and months, so timely observation is necessary to understand and respond to change. By imaging Earth daily, Planet creates a deep historical stack that documents change and enables comparison across time. Democratizing access to satellite imagery empowers journalists, NGOs, governments, and businesses to make better decisions and hold actors accountable. AI can transform raw imagery into a searchable database of objects on Earth, making satellite data far more useful and scalable. A queryable Earth would let users ask direct questions about the planet, similar to how Google made the internet searchable. The combination of satellites and AI can reveal hidden activity, such as illegal fishing vessels that do not transmit AIS beacons.

Data Points: Satellite launches: 21 rocket launches - Number of launches required to build the constellation Satellites in orbit: Over 200 satellites - Current size of Planet’s orbital fleet Ground stations: 31 ground stations - Global network built to receive satellite data Daily imagery volume: 1.5 million 29-megapixel images per day - Amount of Earth imagery downlinked each day Images per location: More than 500 images on average - Average stack of images available for any one location on Earth Time span: Four years - Time since Marshall first announced Planet’s mission at TED Historical reference: 1972 - Year the Apollo 17 blue marble image was taken AI prototype example: Beijing airport plane counts over time - Demonstration of selecting an area and tracking planes across the image stack AI prototype example: Port of Vancouver ship detection - Demonstration of searching for vessels in a selected region

Pivotal Quotes: "you can't fix what you can't see" — Will Marshall: Explaining why up-to-date satellite imagery matters for responding to change on Earth "Mission accomplished." — Will Marshall: Announcing that Planet now images the entire Earth every single day "we're indexing what's on the Earth and making it searchable" — Will Marshall: Summarizing Planet’s long-term vision as a parallel to Google’s indexing of the internet

Implications: Planet’s model could make Earth observation as accessible as web search, accelerating environmental monitoring, accountability, and data-driven decisions across industries and governments.