Document Type : Original Article
Authors
1
Ph.D. Student of Software Engineering, Department of Computer, Rasht Branch, Islamic Azad University, Rasht, Iran.
2
Professor, Computer Engineering Department, University of Guilan , Rasht, Iran, and Department of Computer, Rasht Branch, Islamic Azad University, Rasht, Iran.
3
Assistant Professor, Department of Computer Engineering and Information Technology, Lahijan Branch, Islamic Azad University, Lahijan, Iran.
Abstract
Introduction
Water, as one of the vital and fundamental resources in human life and ecosystem survival, is directly and indirectly affected by pollutants such as heavy metals, pesticides, agricultural toxins, soil erosion, animal waste, and human sewage. This issue has raised significant concerns regarding public health and environmental safety. Rivers, as one of the most important sources of freshwater, are heavily impacted by these changes, emphasizing the necessity of protecting their quality. This text examines methods for assessing water quality and demonstrates how the use of new technologies such as artificial intelligence and the Internet of Things can aid in improving the water quality assessment process. For example, calculating parameter values and converting them into qualitative values can be enhanced using artificial intelligence. According to Zhi et al.'s (2024) research, the determination of all water quality is characterized by traditional measurement methods, which may be time-consuming, expensive, and sometimes in the later stages of evaluation due to the large volume of non-constructible data. It is not reusable. Additionally, this text categorizes challenges and new requirements in this field, indicating that intelligentizing the water quality assessment process through the use of new technologies leads to significant improvements in this area.
Methodology
The methodology of this study consists of several important stages. Initially, by reviewing relevant literature and existing standards in the field of water quality assessment, the necessary foundations for the study were established. Subsequently, through the formation of a Delphi panel and using the snowball sampling method, the opinions of experts and specialists on issues related to water quality assessment were collected and analyzed. In the following steps, by creating a suitable portal for data collection and questionnaire, the required data were gathered from experts and specialists. Then, using the Delphi-fuzzy analysis method, the results obtained from interviews and questionnaires were analyzed by suitable softwares to gain a more precise understanding of the challenges and requirements in water quality assessment. This methodology employs a combined approach, which integrates literature review, expert and specialist participation, questionnaire data collection, and detailed result analysis. After that, we have proposed an IoT-based method for measuring some parameters in river for water quality assessment. The proposed system has been tested in a real environment.
Results and discussion
Results of the research indicated that heavy metal measurement emerged as the most important parameter in water quality assessment, while chlorophyll-a measurement was deemed as the least significant of all. Challenges associated with traditional assessment methods, including the need for expert personnel, costly and time-consuming task, and potential outdatedness of results, were identified. Furthermore, the inadequacy of water sampling conditions underscored the necessity for more online assessment methods. Experts recommended using new technologies such as artificial intelligence and the Internet of Things (IoT) for data collection, storage, processing, and pattern extraction. Based on these findings, a proposed IoT-based system for river water quality assessment was designed, comprising hardware and software components across four layers: perception, network, platform, and application. This system aims to address the limitations of traditional methods while meeting the emerging requirements driven by artificial intelligence.
Conclusions
Water quality assessment, especially the quality of river water, is a highly important process that is traditionally and often manually conducted by some organizations, such as environmental agencies, water and wastewater authorities, and fisheries departments. In this process, experienced experts are dispatched to predetermined locations along the rivers at specific time periods, where some parameters are measured on-site while others are sampled by water and analyzed in laboratories, ultimately resulting in water quality assessment. This paper employed the Delphi-Fuzzy methodology to prioritize influential parameters in water quality and identify existing challenges in the manual water quality assessment process through the assistance of selected experts and professionals via questionnaires and interviews. The analysis of the gathered data revealed that heavy metal assessment parameter holds the highest importance, while chlorophyll-a parameter, holds the least significance in water quality assessment. Furthermore, the challenges in traditional water quality methods, which require sending expert water specialists with suitable equipment to relevant locations at the right time, incurring high costs and time-consuming procedures, were highlighted. Additionally, the results obtained from assessments may become outdated in some cases and lack necessary valuable data. Moreover, water sampling in certain circumstances occurs under unsuitable conditions, necessitating a greater emphasis on online monitoring. Furthermore, experts believed that, considering the aforementioned needs, new technologies such as artificial intelligence and the IoT should be utilized for data collection, storage, processing, and pattern extraction. Based on the conducted research and its outcomes, an IoT-based system for river water quality assessment was proposed. This system comprises a collection of hardware and software in four layers: perception, network, platform, and application. Various sensors in the perception layer are utilized to measure priority parameters. The collected data in the platform layer are processed by different algorithms, and suitable patterns can be extracted using artificial intelligence algorithms and data mining techniques. This proposed system, besides addressing the challenges of traditional methods, possesses the capability to meet the new requirements based on the utilization of artificial intelligence.
Keywords
Main Subjects