Author(s): Ishara Rakith Perera; Joby Boxall; Vanessa Speight; Scott Young; Graeme Moore

Linked Author(s):

Keywords: Data to insights; Efficient data utilization; Information systems; Water utilities

Abstract: Water utility companies collect vast amounts of diverse data types, employing them for regulatory reporting, operational decisions, and strategic planning. The data undergoes intermediary steps from raw collection to insights extraction, which is crucial for informed decision-making. Current practices often involve fragmented data silos, leading to inefficiencies and missed opportunities in data utilization across business applications. This paper identifies key data types in managing water distribution networks (WDNs), categorizing them into static asset, time-series, customer, work management, external, and financial data. A graphical mapping approach is used to connect raw data to applications, enhancing understanding of data usability and identifying underutilized but potent datasets. A critical ranking of intermediary analysis layers highlights high-potential options, including adopting machine learning techniques (especially deep learning) alongside deterministic models for improved decision making. Leveraging less common data types (like asset-specific data) in tandem with traditional datasets shows promise for decision support in WDN management. We expect our attempt to structuralize and understand the usage/ opportunities of data in WDNs can expose new avenues that can be leveraged by advanced analytics tools to further strengthen the water industry's effort to shift from reacting to failures to proactive root cause identification for tackling challenges around WDNs.

DOI: https://doi.org/10.3850/iahr-hic2483430201-406

Year: 2024