<![CDATA[Purpose – Road safety on high-traffic inter-city corridors in Indonesia remains a pressing concern, as drivers receive no real-time hazard notification when approaching zones with statistically elevated crash history. This study develops and evaluates an ESP32-based early warning system that couples GPS-derived positioning with the Equivalent Accident Number (EAN) method to issue graduated audio-visual alerts at road blackspots along the Palur–Semarang bus corridor. Design – EAN quantifies accident severity by weighting fatalities (12), serious injuries (3), minor injuries (1), and property-damage incidents (0.5); segments exceeding the Upper Control Limit (UCL = 170,52) are designated blackspots, with coordinates stored in onboard flash memory. A SIM800L GPRS module transmits positioning data to a web-based fleet monitoring dashboard. Findings – Field evaluation across 10 GPS sampling points yielded mean errors of 0.00033% for latitude (3.7 m) and 0.00005% for longitude (5.0 m), with maximum deviations of 8.9 m and 17.8 m—both within the 800 m geofencing radius. All 10 from 64 validated corridor zones returned EAN values of 199,5–668,5, each exceeding the UCL, with web-platform outputs matching manual calculations exactly. Eight integrated test scenarios confirmed three-tier audio-visual alert delivery at 800 m, 400 m, and 100 m thresholds with zero missed triggers and zero spurious activations. Research implications – These findings provide preliminary evidence for the technical feasibility of EAN-based blackspot intelligence as a driver vigilance aid; however, full-route longitudinal testing across diverse vehicles and network conditions is required before generalised deployment can be recommended. Originality – This study integrates EAN-based crash severity analysis with real-time GPS tracking in an ESP32 system to deliver tiered early warnings for road blackspots. ]]>
Copyrights © 2026
