IoT-based fuzzy logic system for real-time flooded road detection and decision support

Abstract

Flooding remains a critical challenge in urban
environments, often disrupting transportation systems and
endangering drivers who must decide whether to cross inundated
roads without precise knowledge of water depth. Existing IoTbased
flood monitoring systems primarily emphasize data
acquisition and remote alerts but rarely support localized, realtime
decision-making for road safety. This study proposes an IoTbased
flood detection and decision-support system that integrates
Sugeno-type fuzzy logic inference directly into distributed edge
nodes to evaluate flooded road conditions in real time. The system
employs multiple JSN-SR04T ultrasonic sensors and WeMos D1
microcontrollers interconnected via a wireless network to measure
water levels across multiple road segments. Each node
autonomously processes sensor inputs to classify road conditions
into three categories—Safe, Caution, and Not Safe—and displays
results locally via OLED SSD1306 and RYG (Red-Yellow-Green)
LED indicators. Experimental results demonstrate that the
proposed system achieves an average sensor error of 2.2 cm
compared to calibrated measurements and maintains wireless
communication stability up to 25 meters with an average system
response time of 1.19 seconds. Furthermore, the fuzzy inference
outputs from the embedded system closely matched MATLABbased
simulations, validating computational consistency and
inference reliability. The integration of edge-level fuzzy decisionmaking
within distributed IoT nodes represents a key innovation,
enabling autonomous flood assessment with minimal latency. This
study contributes a low-cost, scalable, and intelligent framework
for enhancing road safety and smart city flood resilience,
particularly in developing urban regions prone to recurrent
inundation.