Description
The increasing need for efficient and reliable radiation monitoring in nuclear facilities has driven the development of advanced early warning systems. Conventional Radiation Early Warning Systems (REWS) often rely on manual validation, which can delay critical responses during emergencies. Recent nuclear incidents, such as Fukushima, highlight the importance of real-time, automated monitoring and intelligent evacuation planning to minimize radiation exposure and ensure public safety.
This work introduces GAMOWS, an Artificial Intelligence of Things (AIoT)-based module designed to provide real-time radiation monitoring and intelligent evacuation route recommendations for nuclear facilities. The GAMOWS module integrates several key components and algorithms to achieve these goals, including IoT-based sensors for data acquisition, AI-driven predictive learning for risk assessment, and a modified Dijkstra algorithm for evacuation path planning. Field test simulation results show that the most effective exit route ends at the nearest evacuation path, with a total accumulated dose of 1.355 × 10⁻³ µSv, a travel distance of 19.49 meters, and a required time of 8.77 seconds while running. When the radiation source is hypothesized at the center (reactor core), simulations indicate that the optimal evacuation route from the front door to the nearest exit accumulates a dose of 4.853 × 10⁻³ μSv over a distance of 66.38 meters in 29.8 seconds.
Overall, GAMOWS demonstrates the potential of AIoT-based systems for enhancing radiation monitoring and emergency response in nuclear facilities, where the integration of Artificial Neural Networks (ANN) and intelligent pathfinding enables real-time risk assessment and optimized evacuation, supporting safer and more resilient nuclear operations.
| Technical Track | Student Competition |
|---|
