Predicting the Safest Route During an Earthquake Using AI Planning

Background

After earthquakes, navigating through damaged urban environments becomes perilous due to structural collapses, blocked pathways, and debris. Families face significant challenges in reuniting safely, and emergency response efforts are hindered by unpredictable conditions. These issues highlight the need for a reliable system to predict and verify safe routes.

Objective

• Develop a prototype AI system capable of predicting and verifying safe routes between critical locations like schools, hospitals, workplaces, and homes.
• Reduce travel risks post-earthquake and facilitate faster, safer family reunifications.

Approach

The project employed a combination of advanced methodologies to achieve its goals:

1. Machine Learning Algorithms: Utilized to analyze environmental and structural data for predicting risk scores.
2. Data on Istanbul’s Fatih District: Focused on a high-risk urban area with detailed information on buildings and streets.
3. Risk Score Calculation: Incorporated structural vulnerability, debris probability, and accessibility metrics.
4. Shortest and Safest Path Calculation: Balanced travel time with safety using specialized algorithms.

Results and Impact

The project successfully delivered a prototype AI system that:

• Predicts and Verifies Safe Routes: Ensures accuracy in identifying paths with minimal risks.
• Reduced Travel Risks Post-Earthquake: Provides a safer environment for those navigating after disasters.
• Faster Family Reunifications: Facilitates timely reunification of families separated during the earthquake.
• Improved Emergency Response: Enables efficient resource deployment by emergency teams.

These outcomes demonstrate a transformative potential for urban disaster preparedness.

Future Implications

This innovative AI system offers a scalable and low-cost solution to earthquake-related challenges. It can be deployed globally in other urban areas to bolster emergency response and preparedness, guiding policymaking and further research into integrating AI systems with real-time disaster management strategies.