Automated and dynamic location identification and geofencing based on gps data

Abstract

Aspects of the present disclosure relate to identifying points of interest by generating and storing virtual geofence information that is captured around a physical structure based in part on global positioning system (GPS) data from a plurality of devices that is then processed to identify GPS trajectory and kernel density estimation. Specifically, the techniques include receiving, at the network-based control computer, GPS data from a plurality of devices and grouping the GPS data from the plurality of devices to generate GPS trajectory information for each group of the plurality of devices. Based on the GPS trajectory information, the network-based control computer may calculate kernel density estimation and determine an isoline on a virtual map for the each group of the plurality of devices. By overlaying the isoline data on a geographic coordinate information of a physical structure, the network-based control computer may generate a virtual geofence around the physical structure and store, in a memory, geofence information for the facility.