Term 252

FloodSense Hybrid-Software-First Flash-Flood Risk with Embedded Assist

Project Type: Self-Initiated

Project Description

FloodSense Hybrid presents the final design of a multidisciplinary urban flash flood monitoring and early warning system integrating Computer Engineering (COE), Software Engineering (SWE), Computer Science (ICS), and Industrial & Systems Engineering (ISE) contributions into a unified architecture. The system combines distributed IoT sensing nodes deployed at flood-prone sites with secure cloud-based data processing, AI-driven flood-risk prediction, and optimization-based sensor placement planning. Each site utilizes surface and underground ESP32-based sensor nodes to measure water level, rainfall intensity, blockage percentage, flow rate, and humidity, in addition to external weather API data. Analytical validation confirmed that the sensing subsystem meets the ±1 cm water-level accuracy requirement, supports real-time update constraints, and can theoretically scale to at least 100 distributed nodes while maintaining bandwidth and latency limits. Duty-cycle power modeling demonstrated feasible low-power operation with battery backup capability. The backend infrastructure and dashboard interface was designed to securely manage real-time data exchange and provide interactive monitoring capabilities. The AI module integrates sensor, and forecast inputs to estimate flood risk levels, while the Operations Research model optimizes sensor placement and ensures coverage and cost efficiency under defined deployment constraints. The integrated final design satisfies the defined technical specifications and establishes a scalable, regulation-aligned foundation for future prototype deployment and real-world urban flood early-warning implementation.