DYNAMIC POLYGON GEOFENCING FOR SPATIAL VALIDATION IN CAMPUS ATTENDANCE MONITORING

Abstract

Geofencing systems commonly rely on circular boundaries, which are often insufficient for representing irregular spatial environments such as university campuses. Polygon-based geofencing provides greater spatial flexibility; however, static polygon boundaries become ineffective when user activity areas change over time. This study proposes a dynamic polygonal geofencing approach that adapts its boundary based on user geotagging data. The proposed system constructs a multi-vertex polygon by identifying outermost geotag points relative to a centroid and validating them using proximity-based rules within a predefined distance tolerance. Distance calculations are performed using the Haversine formula. A static polygon derived from benchmark coordinates is used as a reference for evaluation.  Experimental results from six validation scenarios demonstrate that the dynamic geofence progressively improves its spatial alignment with user activity as more geotag data become available. Partial-data testing achieves vertex alignment rates between 33% and 67%, while whole-data testing reaches up to 87% alignment with the six-vertex baseline. Sensitivity analysis further shows that the system remains stable under ±10% tolerance variation. These findings indicate that the proposed method provides a stable and adaptive solution for spatial presence verification in dynamic campus environments.