Evaluating the Performance of an IoT-Integrated Off-Grid Water Treatment System

Many rural communities in Sabah, Malaysia, lack access to centralized water supply systems and rely on selfmanaged solutions for water collection, treatment, and distribution. Without proper monitoring, these systems risk becoming inefficient, unreliable, and difficult to manage. This study presents the design, deployment, and performance evaluation of an IoT-integrated water monitoring system implemented at Kampung Lok Dangkaan, Sabah. The system employs Raspberry Pi controllers, current transformer (CT) sensors, and Node-RED for real-time monitoring, with data transmitted via HTTP and MQTT protocols. Over a 112-day deployment, Site 1 and Site 2 recorded 129,651 and 114,249 messages, respectively, with average daily message counts of 1,165 and 1,023. Missing data rates were 11.34% and 14.37%, while the median transmission interval remained consistent at 60 seconds. The system effectively captured water level recharge–depletion cycles and pump operation events, demonstrating its potential to support sustainable, off-grid water resource management in remote communities.