507 - Designing and Deploying Internet-of-things Enabled Water Diaries to Observe Personal End Uses

Urban water systems are facing supply shortages due to the combined effects of urbanization and the intensification of droughts caused by climate change. To address water shortages, water managers and utilities can employ demand management programs to focus on lowering water demands rather than increasing supply. Characterizing end uses, or the timing and volume of water use at fixtures, can help system managers facilitate water shortage intervention opportunities, design demand management programs, and assess program management effectiveness. While advances in smart meters, which collect data at sub-hourly intervals, have improved methods for characterizing end use, sophisticated disaggregation techniques are needed to classify end use events. Labeled datasets, which identify each trace or flow event, are needed to train end use dissagragation techniques but are not readily available. Labeled datasets necessitate high resolution flow meters at each fixture, which can be obtrusive and expensive. In addition, existing datasets characterize end uses based on fixtures but not individuals, limiting knowledge that can be gained to characterize water use based on individual demographics such as age and gender. This research develops the Electronic Device for Residential Indoor Personal water Sensing (EDRIPS), which is a deployable button array that allows individual residents to quickly and conveniently log water use at fixtures. Individual water users record each use at a fixture by pressing an end-user identification button and a fixture-specific button, which are located on a device near the fixture. The end use event dataset is coupled with smart water meter data, that reports minutely flow data to produce a personal labeled dataset. The EDRIPS system produces individualized end use disaggregation and provides labeled datasets that utilities can use to develop tailored demand management programs. EDRIPS is developed as a prototype of a semi-automated end use data collection system and creates a new dataset characterizing water end uses for individuals.