Application of business intelligence as decision support systems in asset management of water connections

Abstract

When evaluating and applying asset management concepts, water companies can face challenges in enabling targeted recommendation-making due to difficulties in accessing and processing of large volumes of data. These factors can lead to entrance barriers in utilization of breakdown data when assessing network reliability and scope of attainable improvements in asset strategy and maintenance concepts. Proven solutions and contextualized research are not readily available for water companies which, as asset-intensive enterprise relying on physical assets to deliver water to its customers, have a big stake in optimizing its use of data. This project has set out to research whether aspects of recommendation making for asset management at water companies can be aided with application of commonly available and deployable business intelligence tools. To this end, a water company which faces similar challenges has been selected. Evides – a water provider in the region of Rijnmond seeks more data-driven approaches in asset management of water connections. This asset group can be characterized by high volume, high technological heterogeneity and high absolute number of breakdowns as compared to distribution pipes. Together with a vast volume of data, this combination of factors leads to challenges in maintaining a continuous oversight and transparent conversion of performance data into strategic goals and clear service level agreements. The case of Evides inspired a research approach in which application of a custom-made decision support system is evaluated for the process of recommendation making in asset management of water connections. Methodology for this research encompassed for semi-structured interviews with network specialists and managers to obtain information on current asset management goals and the corresponding recommendation-making process for water connections. Thereafter, a thematic analysis was conducted to distill the main themes depicting aspects of interests to network-specialists in charge of producing recommendations and to managers – the decision makers. The type and moments at which performance data is processed and consulted were described and positioned in a managerial decision-making model, together with aspects assessed at each stage. Simple performance indicators were selected to aid the assessments and to connect performance readings with company goals. Findings were thereafter embedded into a purpose-made prototype of a decision support system, utilizing capacity of business intelligence software in creating curated datasets and user-friendly front end. In the last phase of the research, network specialists participated in appraisal of the created tool by completing a series of tasks designed to assess performance of water connections. Surveys were then conducted among participants to evaluate the added value of the created tool in the context of recommendation-making for asset management of water connections at Evides. Results show that, for the case study company, the created tool allows for improvements in accessibility and connectivity of company performance data and can contribute towards greater transparency in goal setting and enabling data-driven recommendation making for asset management of water connections. Performance outliers and policy non-compliers can be localized easier and help company in localizing areas in need of attention. Display of simple performance indicators for connections as per user-selected criteria can in the long run enable more nuances in describing network performance, shifting away from binary descriptions of asset’s performance. In case of Evides, the performance management framework for water connections was discovered as insufficiently defined to allow for assessments of direct benefits as result of application of the designed decision support-system. It is therefore recommended for future research to apply similar methodology for asset groups with well defined performance management standards and to focus on experimental design with higher external validity.