Stays & Travels Metrics

propella.ai leverages de-identified mobile event data ("events") that are geo-coded (given position co-ordinates and a timestamp), aggregated up into weekly data batches by our data provider, and ingested into our Data Lake to power some of our geo-spatial analytics.

Within a given study area, we capture mobile event data to analyse the way in which people utilize and move around public spaces. The raw de- identified mobile device data is pre-processed and aggregated to form distinct measures that are critical to understanding the nature of user interaction with the built environment within that study area.

To help us make better sense of the "sea of mobile device events" that we receive (billions of "pings"), propella.ai has developed a proprietary 'Stays algorithm' which classifies each mobile event into either a 'travel' or a 'stay'.

Travel

A geo-coded event at which a single mobile device was identified as most likely being in motion. A single mobile device can have many travel events registered across different time periods. Travel events help us understand how users move through the geofenced area. Note that although we can associate mobile events with travelling, it is quite difficult to determine the travel mode (walking, cycling, running, driving, etc.) - sometimes we require the location of the mobile event to add context (e.g. a road, footpath, grassed area).

Stay

A geo-coded event where a single mobile device is determined to be stationary (or close to) for a period of one minute or longer. A 'stay' event will be made up of multiple mobile events, where these "pings" all occur within a specific location over a period of time. Stays data is used to inform us about where users/visitors to the geo-fenced study area are stopping ("staying"), which are typically places that people are going to for some intention or purposes (e.g. the picnic area of a park, the public transport stop within a retail precinct).