DroneML

The deployment of state-of-the-art multisensor drone research strategies for archaeological field research results in large multi-resolution and multimodal datasets of significant potential for generating more knowledge about archaeological sites and landscapes. These datasets can encompass a variety of information, such as reflectance maps, vegetation indices, and enhanced visualisations of ground morphology. To identify potential archaeological features, the various data layers must be manually interpreted in search of anomalies—deviations from expected or typical patterns in the data—usually detected through contrasts in the sensor readings. The analysis of such datasets is a time-consuming process of inspecting every individual data layer one by one, marking potential anomalies, and interpreting them in a comparative analysis. This process is further complicated by the fact that archaeological anomalies sometimes only emerge through the fusion of multiple datasets, rather than from the study of individual layers alone. The field of drone remote sensing data analysis would therefore advance fundamentally by the application of computer-aided inspection of these complex datasets. To address this challenge, the 4D Research Lab of the University of Amsterdam, in collaboration with the Netherlands eScience Center, developed CoeusAI, a plugin for the free and open-source geographic information system (GIS) QGIS. The tool has been developed, and a workshop that mainly served as a community launch has been organized, allowing for extensive testing of the software in various contexts.

Although it is too early to report on the eventual impact for the research community, preliminary observations and results are positive. The tool works, and has been applied successfully in the intended context, showing some first positive results (i.e. previously unidentified archaeological anomalies). The reception within the community represented by the workshop participants was positive overall, with more than half of the people willing to use the software on a regular basis. Also, the workshop succeeded in establishing a governance model, involving those present at the workshop. This group is frequently updated on developments and is now involved in case studies that will eventually be published as part of a paper on the project, which is aimed for submission by the end of the year. With this group, we are also exploring the options for follow-up projects, and some ideas are taking shape already.

As such, we can conclude that the project and its results are important, and that the original objectives of the projects have been met.

Main target groups are the international community of the CAA (caa-international.org) and AARG (https://aargonline.com/wp/), but the results have wider relevance for the remote sensing and geophysical prospection community.