How researchers used AI in tandem with drones to find 303 previously uncharted Nazca Lines in Peru, almost doubling the number that had been mapped as of 2020
With drones and A.I., researchers managed to double the number of mysterious geoglyphs in a matter of months.
Context & Ripple Effects
This extends archaeology’s shift toward noninvasive aerial and 3D site surveys, where drones, lidar, radar and mapping tools lower the cost of examining large areas without excavation. It also follows work using neural networks to identify ancient burial sites in satellite imagery, applying machine vision to a different kind of archaeological landscape.
The Nazca result matters because the combined workflow found 303 previously uncharted geoglyphs in months, nearly doubling the count mapped as of 2020. It is a concrete demonstration that AI can turn drone imagery into a faster archaeological-search process rather than merely a documentation tool.
First-order effects
- Researchers and Peruvian heritage stakeholders gain a substantially larger mapped inventory of Nazca geoglyphs, creating a broader base for documentation and study.
- AI-assisted review of drone imagery becomes a proven way to prioritize potential features across the Nazca landscape, reducing reliance on purely manual image inspection.
Second-order effects
- Archaeology teams surveying large or difficult-to-access sites will face stronger incentives to pair aerial data collection with machine-vision screening, building on earlier satellite-based discovery work.
- The larger inventory raises the practical importance of validating, cataloging and managing AI-flagged features; discovery capacity can outpace the expert review needed to turn detections into accepted records.
Third-order effects
- If replicated across sites, archaeology may move from periodic, labor-intensive surveys toward continuously expandable digital inventories built from remote sensing and AI triage.
- The durable constraint shifts from capturing imagery to establishing reliable validation and stewardship workflows, so AI-native sensing is likely to complement rather than replace archaeological expertise.
The trend: AI-native sensing is moving from mapping known terrain to systematically surfacing previously overlooked features in large visual datasets.