DATA-DRIVEN PRESERVATION: A VULNERABILITY AND STRUCTURAL ASSESSMENT OF THE HOSPITAL AT FORT UNION NATIONAL MONUMENT
Open Access
- Author:
- Savaroliya, Mina
- Graduate Program:
- Architectural Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- April 15, 2024
- Committee Members:
- Rebecca Napolitano, Thesis Advisor/Co-Advisor
Ryan Levi Solnosky, Committee Member
Julian Wang, Professor in Charge/Director of Graduate Studies
Thomas E Boothby, Thesis Advisor/Co-Advisor - Keywords:
- Rammed Earth Heritage Structures
Data-Driven Structural Assessment
Fort Union National Monument (FOUN) - Abstract:
- Preserving the integrity of adobe heritage structures presents a significant conservation challenge, necessitating innovative approaches that leverage advanced data-driven techniques to ensure effective intervention strategies tailored to the unique complexities of these architectural treasures. To bridge the gap between extensive datasets and actionable conservation strategies, this project employed machine learning algorithms including Random Forests, Principal Component Analysis and Factor Analysis. These enabled us to develop new ways to analyze and process extensive data and extract meaningful patterns for the development of targeted intervention matrices. These matrices acted as decision-making tools, translating complex patterns and predictions into targeted, prioritized actions for preserving the integrity of adobe buildings. By leveraging big data, we not only developed a new way to predict the most vulnerable aspects of these structures but also formulated data-backed intervention plans that are both proactive and reactive to damage. In this study, rather than relying solely on traditional preservation methods, our approach adapted to the complexities of rammed earth heritage structures. While this work Focuses on the Fort Union National Monument (FOUN) as a case study this research is expected to make significant contributions by advancing our understanding of how we can offer actionable, data-driven solutions for the protection of adobe heritage structures in general.