Agentic AI Hyper-Local Heat Warning System
Agentic AI Hyper-Local Heat Warning System
Anyone who has walked across campus on a very hot day knows that heat can feel very different from one spot to another. This project looks at how HKUST can understand these heat risks more clearly and respond before they become serious. By combining weather data, campus sensors, AI/ML analysis, heat-risk mapping, alerts, and coordinated response workflows, the project aims to support a safer, smarter, and more climate-resilient campus.
What problem is this project trying to address?
Extreme heat is becoming a growing challenge in Hong Kong and it is important to note that heat does not affect every place in the same way. A shaded walkway, an open plaza, a construction area, a sports space, and a steep outdoor route can all feel very different, even on the same day and same timeat the same time. General weather warnings are useful, but they are usually issued at a city-wide or regional level. They do not always show the specific heat conditions experienced by students, staff, visitors, contractors, or outdoor workers in different parts of campus. This project addresses this gap by helping the campus move from broad heat warnings to more local, practical, and coordinated heat-risk management. And it is not only about knowing that the weather is hot. It is also about knowing where the risk is higher, who may be affected, and what action should be taken. Without a shared campus-level system, different teams may respond separately, and useful climate information may not be translated into clear, timely action.
How does the system work?
The system is a smart, campus-specific climate safety platform that acts as an automated "sense-and-respond" network for heat risks:
Sense (IoT & AI): It continuously tracks hyper-local campus temperatures by blending regional weather forecasts with real-time data from on-campus IoT sensors and the university's 3D digital twin.
Warn (Targeted Alerts): When specific spots (like sunny plazas) hit dangerous heat indexes, the system pushes tailored alerts and cooler route recommendations directly to student apps and digital signage.
Act (Agentic AI): Behind the scenes, an intelligent AI orchestrator automatically coordinates and assigns safety tasks to facilities and medical teams to keep the campus safe.
How does this project support our Sustainable Smart Campus as a Living Lab vision?
By using the campus as a real-world testbed for climate resilience, combining data, AI, sensors, and campus operations to improve heat safety and wellbeing.
It supports a smarter and more proactive campus by moving from general weather alerts to local, evidence-based action that can guide both short-term responses and long-term adaptation planning.