Water demand prediction in agriculture, construction industry, and residential sectors using a machine learning model relying on the concept of knowledge management

In contemporary times, the focus of water resource management has shifted from constructing novel water supply systems to the proficient management and utilization of pre-existing systems. Knowledge management is one of the most powerful tools in management science, which is very useful for identifying experimental solutions to this issue. Owing to the fact that machine learning techniques provide ideas for predicting complex phenomena, this study employed the ANFIS model to predict water demand in agriculture, construction, and residential sectors in Mecca Province, Saudi Arabia. Data spanning from 2000 to 2021 was utilized for this purpose. To achieve enough data, the Diz method is utilized for the seasonalization of annual data. The present study assessed and compared the efficacy of water recycling as a means to enhance productivity in the agriculture, construction, and residential sectors in response to water demand management. The findings indicate that the implementation of a water management and recycling strategy can potentially lead to a reduction of 4%, 6%, and 0.8% in water consumption by the agriculture, construction, and residential sectors respectively, by the year 2025. In the absence of management techniques and productivity measures aligned with projected water demand in 2025, the annual consumption levels for the agriculture, construction, and residential sectors are estimated to increase by 20.0, 0.5, and 1.0 MCM, respectively.