Indonesia Trans Java Highway Speed Gap

Abstract

This study examines speed dynamics and surrogate safety on Indonesia's Trans-Java–Cisumdawu corridor using naturalistic dashcam observations of 607 vehicle interactions across eleven segments. Segment-level summaries show substantial heterogeneity in operating speed and speed differentials (speed_gap = v_target − v_source). Overspeeding (>100 km/h) occurred in 45.8% of observations, with notable concentrations on Batang–Semarang, Cisumdawu, and Pemalang–Batang. Robust group comparisons (Welch’s ANOVA with Games–Howell; Kruskal–Wallis) confirmed significant between-segment differences in both speed and speed gaps. A multiple linear model indicated systematic location effects on target-vehicle speed after accounting for traffic density, while a logistic model (AUC=0.876; McFadden’s R²=0.368) showed that overspeeding risk varies markedly by segment even after conditioning on operating speed. Tail-risk indicators – overspeeding prevalence, share of |gap|≥40 km/h, and speed dispersion (CV) – were synthesized into a Speed–Gap Safety Index (SGSI), which prioritized Batang–Semarang, Pejagan–Pemalang, and Semarang as top-tier segments. Time-headway and closing-time diagnostics (short THW/TTC) were consistent with this ordering, indicating frequent entry into closing-risk states. Findings support targeted, segment-specific management – adaptive upper-limit enforcement, speed harmonization/lane discipline, dynamic limits – and provide region-specific inputs for ASEAN NCAP/ADAS scenario calibration (e.g., Δv=20–40+ km/h at THW=0.8–1.2 s). Interpretation is tempered by the observational design, small n on several segments, and video-based measurement, but the results offer an actionable basis for prioritizing enforcement and technology evaluation.