Tainan introduces smart fire simulation system to complete smoke spread prediction within 5 minutes
Quick Look
- The Tainan City Environmental Protection Bureau has built a fire smart simulation and automatic notification service platform.
- Combining AI with HYSPLIT and WRF modes, it can complete smoke dispersion simulation and early warning of the impact area in as fast as 5 minutes, automatically compare sensitive places and air quality data, remind the public to protect themselves, and shorten response time.
AI-generated summary
Why It Matters
Tainan City has a large number of factories and the weather is dry in autumn and winter. Fires occasionally occur in factories, warehouses, etc. Traditional fire response requires manual collection of data and setting of simulation conditions, which takes at least 10 minutes to produce an impact area prediction.
Tainan has introduced smart fire simulation to provide real-time warning of the impact of smoke flow. (Provided by Nanshi Environmental Protection Bureau)
The Tainan City Environmental Protection Bureau built the "Tainan City Fire Smart Simulation and Automatic Notification Service Platform" this year. In the past, when a fire broke out, personnel had to collect data, set simulation conditions, and then generate the likely affected areas, which took at least 10 minutes. Now, with the assistance of AI and automation systems, smoke flow diffusion simulations can be completed in as fast as 5 minutes, the scope of impact can be determined, and responses can be initiated immediately to remind people in the affected areas to take precautions and gain valuable response time.
Tainan has a large number of factories and the weather is dry in autumn and winter. Fires occasionally occur in factories, warehouses and other places. The Environmental Protection Bureau interfaces with the fire bureau’s real-time fire information through the system to obtain fire location, time and fire brigade dispatch information. When it is judged to be a specific type of fire and more than 10 squads are dispatched, pollution diffusion simulation will be automatically started to reduce manual operation time.
Please read on...
The system uses the HYSPLIT airflow trajectory model approved by the Ministry of Environment, combined with the WRF mesoscale meteorological grid model with a resolution of up to 3 kilometers, to predict the range and trajectory of pollution diffusion, and simultaneously compares sensitive places such as nearby schools, hospitals, and nursing homes, as well as air quality measuring stations and micro-sensor data. If an abnormality is found, it will be alerted immediately.
The Environmental Protection Agency stated that the biggest change after the introduction of AI smart tools is "time saving and quick response", allowing personnel to spend time on judgment and response, and immediately reminding people in downwind areas to close doors and windows and take precautions to reduce the impact of fire smoke on life and health.
Open Questions
- How much does it cost to build and maintain the system?
- Will it be expanded to other counties and cities in the future?
- How accurate is the system in extreme weather conditions?





