N.M. Leyva-Munguia, R. Maciel
Universidad de Guadalajara,
Mexico
Keywords: drinking water quality, internet of things, systematic literature review, sensors
Summary:
Objective: This study systematically evaluates the state of the art in Internet of Things (IoT) based systems for drinking water quality monitoring, focusing on their potential to mitigate inequitable access to safe drinking water across rural and low-income regions. As the availability and access to potable water remain a primary global challenge, IoT solutions are emerging as a promising approach to bridging technological and socioeconomic gaps in water management. Methodology: A systematic literature review was conducted following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) methodology. Searches were performed in the Web of Science, Scopus, IEEE Xplore, and ProQuest databases using defined inclusion criteria. From an initial set of 2,977 publications, multi-stage screening removed duplicates and irrelevant studies. Subsequently, 30 studies were assessed for eligibility, evaluated against five quality indicators: clarity of measured parameters, detailed system description, completeness of results and limitations, peer-review status, and appropriate integration of Machine Learning and Deep Learning techniques as a complement to the IoT system. This process yielded 18 articles that fulfilled all requirements for the final synthesis. Results: The findings confirm the viability and accessibility of low-cost IoT systems, dominated by open-source platforms like Arduino and ESP microcontrollers. These systems enable cost-effective, real-time monitoring of pH, turbidity, TDS, and temperature, these being key physicochemical measured parameters. A significant part of the synthesized studies (16 out of 18) evaluated their measurements against international standards, confirming their direct applicability of these technologies to public health monitoring. The integration of Artificial Intelligence and Machine Learning tools represents a significant advancement, with classifiers successfully applied to water potability prediction and anomaly detection. Furthermore, LoRa and NB-IoT connectivity enhance communication reliability in remote or infrastructure-limited areas. A notable breakthrough was identified in the development of an ultra-sensitive FET hybrid sensor for arsenic detection, representing a major step toward trace contaminant monitoring. Conclusions: IoT-based water quality monitoring systems represent a paradigm shift in sustainable resource management, offering real-time data acquisition, portability, and low-cost deployment. However, significant challenges persist: the lack of robust, low-cost sensors for chemical and biological contaminants, limited studies on long-term operational stability, and insufficient evaluation of energy consumption and sensor durability. Moreover, data security and user-centered design remain unexplored dimensions for scalability and social adoption. Addressing these gaps is key to transitioning from experimental prototypes to fully autonomous, resilient, and community-driven monitoring networks, particularly in Latin American contexts where equitable access to safe water remains an urgent priority.