Page 685 - AI for Good Innovate for Impact

AI for Good Innovate for Impact

archaeological zones. These structured datasets, derived from both remote sensing analysis
and field validation, support archaeological interpretation, spatial modeling, and decision-
making—reinforcing the approach’s value in both research and heritage management.

Use Case Status: Pilot completed. The use case has moved beyond the initial research phase cities 4.8: Smart home/
and was successfully implemented and tested at the Saruq Al-Hadid archaeological site in
Dubai, UAE, serving as a full-scale pilot to validate the proposed methodology.

Partners: Dubai Culture & Arts Authority, Khalifa University�

2�2 Benefits of the use case

The use case delivers significant and multifaceted benefits across education, research,
conservation, and public sector innovation. It advances academic knowledge and supports
training in archaeological and remote sensing techniques, contributing to quality education
and lifelong learning. As a non-invasive innovation, it enabled the detection of archaeological
anomalies at the Saruq Al-Hadid site without physical excavation—preserving site integrity and
ensuring sustainable conservation practices. The approach reduced fieldwork time and cost,
enhanced operational efficiency, and guided excavations using high-resolution maps and
3D digital models. Strategically, it aligns with national goals in cultural sustainability, scientific
research, and smart urban development. The project also encouraged interdisciplinary
collaboration among governmental authorities, academic institutions, and data providers,
fostering effective partnerships and knowledge exchange. It strengthened the role of cultural
authorities as public sector innovation leaders and introduced a framework for archaeology–AI
collaboration. Marking a shift from traditional manual surveys to AI-driven techniques rarely
adopted in the region, the initiative is the first in the world to fully integrate satellite remote
sensing, machine learning, and geophysical data (GPR and magnetometry) for archaeological
discovery. This service and policy innovation introduced AI-powered site detection as a new
exploration method, creatively combining diverse datasets into a unified, predictive workflow.
The methodology is scalable and adaptable, supporting replication by other emirates, cultural
bodies, or global heritage institutions, and applicable across sectors such as environment,
tourism, and smart city planning. It is ideal for heritage sites under urban infrastructure or in
fragile ecosystems and provides a blueprint for AI-powered cultural resource management.
The project enables efficient assessments without disrupting communities or delaying
development, supports academic publishing and public engagement, and contributes to
preserving archaeological heritage for future generations. Ultimately, it exemplifies how AI can
enhance cultural preservation, marking not only a technological breakthrough but a milestone
in global heritage conservation.

2�3 Future Work

Model Enhancements and Generalization:

Future improvements will focus on integrating advanced neural network and backpropagation
algorithms to boost detection and classification accuracy. Model generalizability will be
enhanced through the adoption of transfer learning and data augmentation techniques
(e.g., rotation, scaling), enabling performance adaptation to other arid regions and diverse
archaeological contexts. Establishing a framework for long-term multimodal investigations will

649