Using Data Analytics and Artificial Intelligence-based techniques to promote water-conscious behaviour through personalised recommendations

University of Technology Sydney. Faculty of Engineering and Information Technology.Water scarcity is one of the largest global risks affecting many countries around the globe. Many urban water utilities are facing the challenge of ensuring a sufficient water supply during prolonged drought. Avoiding low pressure during the hours of peak demand is another challenge that is also troubling metropolitan water utilities. To overcome these challenges, digital water meters are considered a potential solution. However, the scope to fully utilise the data collected from digital water meters using data analytics and artificial intelligence-based techniques is currently unrealised. This research investigates the applications of data analytics and artificial intelligence-based techniques to promote water-conscious behaviour through personalized recommendations. First, a systematic literature review was conducted to determine the state-of- the-art in digital water metering data analytics. Second, to achieve personalisation at the household level, an advanced household profiling approach was proposed. Third, to understand the water consumption patterns of different groups of households, a clustering approach was proposed. Lastly, a generic model was proposed to integrate recommender systems in behaviour change interventions; a prototype of a recommender system was developed based on the proposed model; and people’s attitudes towards the personalised recommendations were identified through a survey questionnaire