Real-Time Fire Risk Assessment in Server Rooms Using Hybrid Threshold and Fuzzy Logic on Raspberry Pi

Abstract

The server room is a critical area for the storage and operation of server-related machines. In this environment, it is essential to protect data and systems from unwanted events. Significant increases in temperature or humidity can cause severe damage to equipment, system failures, and even fire risks due to sparks from transmission cables. Temperature and humidity readings with average values of 29.83°C and 45.5%, respectively, and low standard deviations, indicating reliable measurements. with an average delay of approximately 2.5 milliseconds. The system continuously monitors temperature, humidity, and flame presence to assess risk levels accurately. Threshold methods provide immediate alerts when sensor readings exceed predefined limits, while fuzzy logic enables nuanced risk classification based on combined sensor inputs. Experimental results demonstrate the system's effectiveness in early detection of hazardous conditions via visual and auditory alarms, thereby enhancing the protection of critical server infrastructure.