Ancient Aliens Need Not Apply: What to Know About How AI, ML, and LiDAR are Discovering Amazing Prehistoric Civilizations In Places We Never Dreamed

Recent research at the intersection of artificial intelligence (AI), machine learning (ML), and remote sensing—including LiDAR—has transformed the discovery and study of ancient civilizations, particularly in Central America and Mexico.

Recent Advances and Findings

Central America and Mexico: The Maya Example

LiDAR and Machine Learning for Urban Mapping:
LiDAR technology, combined with advanced ML algorithms, has enabled archaeologists to map vast areas of dense jungle and uncover extensive networks of Maya cities, roads, and infrastructure. These findings have challenged previous assumptions that the Maya civilization was composed of small, isolated villages. Instead, researchers have revealed a complex, interconnected society with advanced trade routes and infrastructure, including causeways up to 131 feet wide, which functioned as early superhighways connecting cities across long distances12.
Recent Discoveries:
In 2024, a massive Maya city—Valeriana, comparable in size to Edinburgh and among the largest in ancient Latin America—was discovered in the jungles of Campeche, Mexico. This discovery was made possible by analyzing LiDAR data, which revealed pyramids, sports fields, causeways, and amphitheaters hidden beneath the dense canopy2.
Machine Learning-Ready Datasets:
Open datasets, such as those for the Chactún region (a major Maya urban center in the Yucatán), provide high-resolution LiDAR, Sentinel-1 SAR, and Sentinel-2 optical satellite data, all annotated for specific Maya structures (buildings, platforms, and aguadas). These datasets are designed for use with convolutional neural networks (CNNs) to automate object recognition and localization, supporting large-scale archaeological surveys and competitions3.

Global Examples and Methodologies

AI-Assisted Site Detection:
AI and ML models are being used globally to process remote sensing data (LiDAR, satellite, UAV imagery) for archaeological site detection. For example, projects in the central Andean cordillera and Mesopotamian flood plains have demonstrated the effectiveness of AI in identifying hidden structures and landscape features45.
Citizen Science and Human-AI Collaboration:
Projects like the Dutch ‘Heritage Quest’ combine LiDAR, AI, and public participation to map thousands of potential archaeological sites with high accuracy. Volunteers help refine AI models by identifying subtle patterns that algorithms might miss, improving both detection rates and model performance6.
Automated Data Processing:
Recent research highlights workflows that use ML for filtering vegetation from LiDAR data and detecting archaeological features. These workflows are designed to be transferable across different environments, making them adaptable for a wide range of archaeological contexts75.

Key Technologies and Approaches

Technology/Method

Application in Archaeology

Impact/Example

LiDAR

Mapping under dense vegetation

Maya cities, Valeriana12

Machine Learning (ML)

Automated site detection, feature mapping

Chactún dataset, Maya roads3

Satellite/UAV Imagery

Large-scale landscape surveys

Andean, Mesopotamian studies4

Citizen Science

Human verification, model training

Heritage Quest (Netherlands)6

Summary

The integration of AI, ML, and remote sensing—especially LiDAR—has revolutionized the discovery of ancient civilizations. In Central America and Mexico, these technologies have led to groundbreaking findings about the scale and complexity of the Maya civilization, uncovering vast urban networks and infrastructure previously hidden by jungle. Globally, AI and ML are accelerating archaeological surveys, enabling non-invasive exploration, and fostering collaboration between archaeologists, computer scientists, and the public.

I hope you enjoyed this Introduction to how AI, ML, and LiDAR are opening up new amazing Archaeology Discoveries!

Chris Coffey

BlueLensAI

21 June 2025