Conference Agenda

Archaeological sites across the globe are facing significant threats and heritage managers are under increasing pressure to monitor and preserve these sites. Since 2015, the EAMENA project has documented more than 200,000 archaeological sites and the disturbances and threats affecting them across the Middle East and North Africa (MENA) region, using a combination of remote sensing, digitization, and fieldwork methodologies. The large number of sites and their often remote or otherwise difficult to access locations makes consistent and regular monitoring of these sites for disturbances and threats a daunting task. Combined with the increasing frequency and severity of threats to archaeological sites, the need to develop novel tools and methods that can rapidly monitor the changes at and around archaeological sites and provide accurate and consistent monitoring has never been more urgent.

In this workshop, we will introduce the EAMENA Machine Learning Automated Change Detection tool (EAMENA MLACD). This newly-developed online tool uses bespoke machine learning algorithms to process sequential satellite images and create land classification maps to detect and identify disturbances and threats in the vicinity of known archaeological sites for the purposes of heritage monitoring and preservation. The tool is developed in Google Earth Engine with a user-friendly interface and workflow, which requires only basic knowledge of GIS and remote sensing making it a powerful tool. Initial testing and validation of results from the EAMENA MLACD in a case study in Bani Walid, Libya, demonstrate how it can be used to rapidly identify disturbances and potential threats to heritage sites, and increase the speed and efficiency of monitoring activities undertaken by heritage professionals.

Supported by the British Council’s Cultural Protection Fund, the EAMENA project team has already provided training to over twenty heritage professionals in Libya and Algeria on how to use and adapt the EAMENA MLACD tool for the purpose of heritage preservation in their countries.

References

Mahmoud, Ahmed Mutasim Abdalla and Sheldrick, Nichole and Ahmed, Muftah, A Novel Machine Learning Automated Change Detection Tool for Monitoring Disturbances and Threats to Archaeological Sites. Available at SSRN: https://ssrn.com/abstract=4914336

Rayne, L., et al. 2020. “Detecting change at archaeological sites in North Africa using open-source satellite imagery.” Remote Sensing 12(22):3694. https://doi.org/10.3390/rs12223694

Deep Learning applications in optical based imagery are transforming archaeological prospections by shifting the research landscape from manual analysis of large-scale territorial data to automated methods. These technologies provide substantial advantages for the exploration and preservation of archaeological sites. However, deep learning, despite its potential, has yet to establish widely accepted best practices for detecting subsoil anthropogenic or environmental features, which vary widely in size, shape, and geographical context. The research undertaken at CCHT is addressing this issue by combining machine learning with remote sensing technologies to identify a broad spectrum of subsurface features, providing a powerful tool for archaeological research.

Ongoing work focuses on the development of a comprehensive framework for detecting archaeological features using multispectral satellite data. Through practical case studies, partially from the Cultural Landscapes Scanner (CLS) project is presentation aims to provide fresh insights into the challenges of identifying subsurface features in remote sensing imagery. The presentation specifically will address the issues of limited feature visibility and the scarcity of publicly available datasets by creating a multitemporal dataset of multispectral images and employing semantic segmentation across different seasons to classify palaeochannels using Sentinel-2 time-series imagery. The framework is designed to ensure the continuity of these features despite seasonal variations, reduce background interference that affects visibility, and assess performance metrics under various environmental conditions to determine the most favourable circumstances for feature detection. The experiments seek to lay the groundwork for applying these methods to the detection of smaller archaeological traces in higher-resolution imagery, where detection is more challenging due to the limited size of available datasets.

The world is constantly changing and becoming more complex in every aspect. In this complex environment, cultural heritage has become increasingly important. Heritage sites play a crucial role in global economic and cultural activities. Today's world seeks to monitor, assess, and preserve cultural heritage assets as part of sustainable development. They are often complex structures with large spatial scales, which makes visualization and analysis challenging. Traditional methods may not fully represent these assets, hindering detailed analysis and decision-making. Novel monitoring and detection systems are required to streamline and simplify the auditing process of CH assets. One of the recent advancements in digital technology is AI-based automation. The approach involves developing AI-powered algorithms in the space supported monitoring process and advanced deep learning models to detect damages and monitor digital replicas of cultural heritage assets on a 3D/4D virtual representation of the Earth's surface within a 3D+ virtual environment to address asset degradation. The method consists of exploratory data acquisition using Unmanned Aerial Vehicles (UAV), creating Digital Twin, developing virtual globe based portal monitoring system. This enhances the precision and effectiveness of detecting damages and changes, ultimately creating a 3D+ virtual space. Early detection of damage in cultural heritage assets is crucial for preventive conservation measures. The model enable timely identification of potential threats to cultural heritage assets, allowing for proactive intervention and preventive measures and facilitates the documentation, monitoring, and preservation of Cultural Heritage.

Aerospace engineering and science’s contribution to multiple game-changing technologies that have become essential to society is difficult to underestimate: from LASIK eye-tracking technology in eye surgery to ACTIS from the Apollo mission, which made CAT scans possible, to insulin pumps, scratch-resistant lenses, and many others.

NASA and MOXY pioneered atomic oxygen (AO) technology, which may become a game-changing contribution of aerospace science to society and cultural heritage (CH). AO has a unique potential for non-contact and solvent-free cleaning of invaluable but extremely sensitive materials.

The presentation will share the concept of the tailored AO process and recent experimental results of the MOXY Horizon Europe project (moxyproject.eu) at its midterm point (2022-2026). MOXY has developed a groundbreaking non-contact cleaning technology that uses RF plasma-generated AO at atmospheric pressure to remove problematic carbon-based contaminants such as environmental pollution, fire-born soot, fatty acid exudates, bio growth, and sebum from challenging cultural heritage surfaces in an entirely non-mechanical, solvent and liquid-free manner, without health concerns or environmental residues or waste.

AO will empower CH conservators in diverse specializations with a new technology that resonates with the urgent need to refine cleaning methodologies and gain treatment readiness for fire damage while embracing climate resilience in our studio practice.

The project is part of a broader strategy for Horizon Europe and the EC to fund and advance green methodologies for cultural heritage conservation, achieve the aims of the European Green Deal, and respond to UN Sustainable Development Goals (SDG 11.4). Three Horizon Europe projects, MOXY (Ghent University), GoGreen (University of Amsterdam), and GreenArt (University of Florence), have formed a “Green Cluster” to find synergy in green methodologies for CH, training, and advocacy. On September 4, the three projects held a joint symposium in Vilnius, Lithuania, on Green and Sustainable Approaches to Cleaning Sensitive Works of Art.