An article recently posted on the YaleNews website demonstrated the innovative work of Holly Rushmeier, the John C. Malone Professor of Computer Science at Yale School of Engineering & Applied Science. Her team used artificial intelligence (AI) to reconstruct the ancient city of Dura-Europos in present-day Syria and assess damages caused by forest fires in Algeria. The research aimed to show the potential of AI to revolutionize fields like archaeology and environmental conservation.
Study: How AI can reveal new understandings of the past — and the future. Image Credit: JOURNEY STUDIO7/Shutterstock.com
Background
AI is now used in many fields, such as education, healthcare, retail, and business, to analyze complex data and support faster decision-making. In archaeology, AI helps reconstruct ancient sites and artifacts such as buildings and entire cities, providing detailed insights into historical civilizations. This technology enables the preservation of cultural heritage through digital reconstructions that can be shared with both the public and researchers.
In environmental conservation, AI helps assess damage from natural disasters like forest fires, floods, earthquakes, avalanches, and hurricanes. By combining machine learning algorithms with satellite imaging, researchers can quickly analyze remote sensing data.
This enables them to predict environmental impacts, track changes in ecosystems, and improve disaster response strategies. These AI-driven insights support better conservation practices and help address the effects of climate change.
About the Study
Rushmeier and her team focused on two key projects. The first involved using AI to reconstruct Dura-Europos, an ancient city founded in 300 B.C.E. and abandoned in the third century C.E. The city was known for its cultural and linguistic diversity, but much information about it has been lost over time.
The team used AI to create three-dimensional (3D) models of the city’s buildings, extracting key features from historical photos for geometric modeling. They also collected data from excavation reports and museum collections to integrate into a "linked open data" system, forming a knowledge graph for future research through platforms like Wikidata. This system allows for the integration of fragmented data from various sources, providing a comprehensive understanding of the city's history and culture.
The second project focused on assessing land damage caused by forest fires in northern Algeria. Since on-site data is hard to gather, the researchers relied on satellite data from European and U.S. satellites. Furthermore, Ph.D. student Nadia Zikiou compared convolutional neural networks (CNNs) and support vector machines (SVMs) to predict wildfire damage and track recovery.
The aim was to find the best methods and wavelengths for assessing vegetation damage using satellite data. This project also explored the use of hyperspectral data, which provided valuable insights into vegetation recovery and land restoration.
Key Outcomes
In the Dura-Europos project, AI models successfully reconstructed 3D images of several buildings, offering a clearer view of the city’s layout and architecture. Integrating data into a linked open data system has improved access to information about the site, preserving historical knowledge and supporting future research and education.
The Algerian forest fire project provided key insights into the effectiveness of machine learning methods. The comparison between CNNs and SVMs showed that CNNs proved more accurate than SVMs in predicting wildfire damage.
The study also identified the best wavelengths for analyzing satellite data, enabling more precise assessments of vegetation damage and recovery. Additionally, hyperspectral data was particularly effective in tracking vegetation recovery, offering valuable insights for land restoration.
Applications
This study has several applications. The 3D reconstruction of Dura-Europos offers detailed insights into the city’s layout, benefiting archaeologists and historians. It also serves as an educational tool for the public. The linked open data system supports further research and promotes the integration of fragmented data in other archaeological projects.
In environmental science, the AI methods developed for assessing wildfire damage can be applied to other regions and natural disasters. Accurate monitoring of ecological damage is crucial for resource management and recovery. These tools can help governments and organizations improve land recovery strategies and respond more effectively to environmental challenges.
Conclusion
Holly Rushmeier’s innovative application of AI in archaeology and environmental science demonstrated the transformative potential of this technology. Her work, which included reconstructing the ancient city of Dura-Europos and assessing the impact of forest fires in Algeria, provided valuable insights and practical applications. The study not only enhances the understanding of historical sites and natural disasters but also offers useful tools for future studies and resource management.
Journal Reference
Weir, W. How AI can reveal new understandings of the past — and the future. YaleNews Website, 2024. https://news.yale.edu/2024/09/11/how-ai-can-reveal-new-understandings-past-and-future
Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.