Harnessing IoT and advanced analytics for sustainable water quality management

Abstract

• Real-time IoT-enabled system improves water quality monitoring in the University of Ilorin. • Advanced analytics identify decreasing trends in Electrical Conductivity (EC) and Total Dissolved Solids (TDS). • Canadian Water Quality Index (CCME WQI) categorizes water quality as ’Excellent’ with a score of 94.42 %. • Principal Component Analysis (PCA) reveals EC and TDS as primary factors influencing water quality variability. • Solar-powered IoT system offers sustainable, continuous water quality monitoring for public health safety. Access to safe and clean water remains challenging in resource-constrained environments, where conventional laboratory-based assessments suffer from delayed feedback and limited sampling. This study developed and validated a solar-powered Internet of Things (IoT)–enabled real-time water quality monitoring framework, integrating sensor networks with cloud-based analytics. A weatherproof sensor system was deployed at the University of Ilorin Water Treatment Plant, collecting 100 time-stamped observations over 13 days. The system measured six key physicochemical parameters—pH, turbidity, temperature, oxidation–reduction potential (ORP), electrical conductivity (EC), and total dissolved solids (TDS). The Canadian Council of Ministers of the Environment Water Quality Index (CCME-WQI) yielded a score of 94.42 %, classifying the water as “ excellent ” . Pearson correlation revealed strong relationships between turbidity and temperature ( r = 0.59) and temperature and EC ( r = 0.43). Trend analysis using the Mann–Kendall test showed significant increases in turbidity ( τ = 0.339) and EC ( τ = 0.222), while pH declined ( τ = –0.383). Corresponding Sen’s slopes confirmed gradual daily changes. OLS regression supported turbidity’s upward trend ( β = 0.0261, R 2 = 0.202). Principal Component Analysis (PCA) reduced the dataset to three components, with the first two explaining 57.26 % of the total variance. PC1 was associated with TDS, EC, and turbidity; PC2 with pH and ORP; and PC3 with temperature. These results demonstrate the system’s capability for automated water quality assessment. Future work should expand temporal coverage, adopt predictive modeling, and extend deployments across multiple sites for broader impact.