Inferring dwelling occupancy patterns from high temporal resolution water metering data
Topics:
Keywords: Water consumption, smart metering, dwelling occupancy patterns, tourism.
Abstract Type: Paper Abstract
Authors:
Owen Hibbert, Leeds Institute for Data Analytics (LIDA), University of Leeds
Andy Newing, School of Geography, University of Leeds
,
,
,
,
,
,
,
,
Abstract
The rollout of advanced metering infrastructure in the residential water supply sector affords tremendous benefits in driving water-use efficiencies, accurate billing and network management (e.g. leak detection). These data, which are collected at high-temporal resolution at a dwelling level, could offer a non-intrusive means of inferring dwelling water use characteristics and occupancy patterns (water is typically only consumed when householders are present). These insights could have a range of benefits dependent upon the spatio-temporal scale and the intended application. Our interest is in using these data to identify dwellings that have inferred occupancy patterns associated with tourism, such as second homes or short-term holiday rentals. Specifically, we report on research that develops and recommends analytic tools suitable for extracting dwelling-level occupancy features associated with tourism from household level-water consumption data. Those features include periods of unoccupancy, seasonal occupancy patterns associated with second homes and an absence of consumption trends that may be associated with habitual residential routines. We illustrate these techniques using household level data for a sample of households in Devon and Cornwall, South West England. Our data partner, an English regional water supplier, has provided these data. These data capture consumption pre-Covid, and during Covid ‘lockdown’ and ‘staycation’ periods, enabling us to extract a range of interesting property-level occupancy trends. We reflect on the potential benefits of these insights in inferring dwelling and neighbourhood type and consider the wider-reuse value of these data in generating area-based population statistics.
Inferring dwelling occupancy patterns from high temporal resolution water metering data
Category
Paper Abstract