This study tested a new way to plan disaster shelters in oil and gas cities using artificial intelligence . Researchers used an “artificial bee colony” algorithm to choose shelter locations, size, and resources based on risk, population, and transportation. Results showed improved disaster resilience (average score 0.64) and faster evacuation times, with most residents reaching shelters within about 10 minutes at 4 km distance. AI-based shelters also improved resource use, renewable energy use, and community participation compared to traditional planning. The study suggests AI tools can help planners design safer, more resilient shelters for high-risk industrial cities

Intervention Area: