Archaeological surveys conducted through the inspection of high-resolution satellite imagery promise to transform how archaeologists conduct large-scale regional and supra-regional research. However, conducting manual surveys of satellite imagery is labour- and time-intensive, and low target prevalence substantially increases the likelihood of miss-errors (false negatives). In this article, the authors compare the results of an imagery survey conducted using artificial intelligence computer vision techniques (Convolutional Neural Networks) to a survey conducted manually by a team of experts through the Geo-PACHA platform (for further details of the project, see Wernke et al. 2023). Results suggest that future surveys may benefit from a hybrid approach—combining manual and automated methods—to conduct an AI-assisted survey and improve data completeness and robustness.

Imagery-based survey is capable of producing archaeological datasets that complement those collected through field-based survey methods, widening the scope of analysis beyond regions. The Geospatial Platform for Andean Culture, History and Archaeology (GeoPACHA) enables systematic registry of imagery survey data through a ‘federated’ approach. Using GeoPACHA, teams pursue problem-specific research questions through a common data schema and interface that allows for inter-project comparisons, analyses and syntheses. The authors present an overview of the platform's rationale and functionality, as well as a summary of results from the first survey campaign, which was carried out by six projects distributed across the central Andes, five of which are represented here.

By mapping sites at large scales using remotely sensed data, archaeologists can generate unique insights into long-term demographic trends, inter-regional social networks, and human adaptations in the past. Remote sensing surveys complement field-based approaches, and their reach can be especially great when combined with deep learning and computer vision techniques. However, conventional supervised deep learning methods face challenges in annotating fine-grained archaeological features at scale. In addition, while recent vision foundation models have shown remarkable success in learning large-scale remote sensing data with minimal annotations, most off-the-shelf solutions are designed for RGB images rather than multi-spectral satellite imagery, such as the 8-band data used in our study. In this paper, we introduce DeepAndes, a transformer-based vision foundation model trained on three million multi-spectral satellite images, specifically tailored for Andean archaeology. DeepAndes incorporates a customized DINOv2 self-supervised learning algorithm optimized for 8-band multi-spectral imagery, marking the first foundation model designed explicitly for the Andes region. We evaluate its image understanding performance through imbalanced image classification, image instance retrieval, and pixel-level semantic segmentation tasks. Our experiments show that DeepAndes achieves superior F1 scores, mean average precision, and Dice scores in few-shot learning scenarios, significantly outperforming models trained from scratch or pre-trained on smaller datasets. This underscores the effectiveness of large-scale self-supervised pre-training in archaeological remote sensing. Codes will be available on https://github.com/geopacha/DeepAndes.

Fog oases (lomas) present pockets of verdant vegetation within the arid coastal desert of Andean South America and archaeological excavation within some of the oases has revealed a long history of human exploitation of these landscapes. Yet lomas settlements are under-represented in archaeological datasets due to their tendency to be located in remote inter-valley areas. Here, the authors employ satellite imagery survey to map the locations of anthropogenic surface features along the central Peruvian coast. They observe two categories of archaeological features, large corrals and clustered structures, and document a concentration of settlement features within lomas landscapes that suggests a pre-Hispanic preference for both short- and long-term occupation of these verdant oases.

The monumental scale agricultural infrastructure systems built by Andean peoples during pre-Hispanic times have enabled intensive agriculture in the high-relief, arid/semi-arid landscape of the Southern Peruvian Andes. Large tracts of these labor-intensive systems have been abandoned, however, owing in large measure to a range of demographic, economic, and political crises precipitated by the Spanish invasion of the 16th century CE. This research seeks to better understand the dynamics of agricultural intensification and deintensification in the Andes by inventorying through the semantic segmentation of active and abandoned agricultural fields in satellite imagery across approximately 77,000 km² of the Southern Peruvian Highlands. While manual digitization of agricultural fields in satellite imagery is time-consuming and labor-intensive, deep learning-based semantic segmentation makes it possible to map and classify en masse Andean agricultural infrastructure. Using high resolution satellite imagery, training and validation data were manually produced in distributed sample areas and were used to transfer-train a convolutional neural network for semantic segmentation. The resulting dataset was compared to manual surveys of the region and results suggest that deep learning can generate larger and more accurate datasets than those generated by hand.

El sitio de Huari, centro una vasta formación política expansionista, se cuenta entre las ciudades más grandes de Sudamérica precolombina. La ciudad existió durante aproximadamente cuatrocientos años durante el Horizonte Medio (600-1000 EC). Señalado como uno de los primeros centros de la civilización andina por Cieza de León en el siglo XVI, e investigado por primera vez por Julio C. Tello en 1931, se ha establecido su importancia a través de casi un siglo de investigación arqueológica. Sin embargo, la misma escala del sitio ha complicado la comprensión de su organización espacial. Hasta la fecha, solo existen planos esquemáticos de la ciudad sobre la base de fotos aéreas y sectores mejores documentados a través de levantamiento cartográfico en el campo. Para comprender mejor el sitio, en 2017 el Museo Real de Ontario y la Universidad de Vanderbilt organizaron un proyecto conjunto de registro espacial que cubrió la porción de más de 2 km2 del sitio que contiene arquitectura permanente. Este artículo analiza la historia de los proyectos de levantamiento cartográfico de Huari antes de 2017, así como la metodología que usamos para adquirir y procesar ortomosaicos de Huari en gran parte mediante el uso de drones (UAV). Nuestro proyecto produjo imágenes en la mayoría del asentamiento con una resolución de hasta ~ 3 cm, disponible hoy para exploración y descarga en www.huarimappingproject. org. La plataforma se puede actualizar a medida que haya más imágenes disponibles. Los ortomosaicos, así como los modelos derivados de ellos, se pueden utilizar para respaldar la investigación futura y los programas de preservación del patrimonio.