Bottleneck analysis fuses two complementary data sources: TomTom probe vehicle GPS data at link level, and SCATS traffic signal controller data at intersection level. This dual-layer fusion distinguishes between link-demand bottlenecks and signal-capacity bottlenecks — and detects queue spillback where a saturated intersection is blocking upstream links.
TomTom Probe Data — Links
Anonymised GPS probe traces aggregated per corridor segment and hour. Travel Time Index (TTI) measures congestion relative to free-flow speed. Buffer Time Index (BTI) captures travel time unreliability from P85 spread.
TTI = free-flow speed / current speed
BTI = (P85 TT − median TT) / median TT
SCATS Signal Data — Intersections
SCATS detector counts provide degree of saturation (DoS) per intersection per hour. Control delay is derived via Webster's simplified formula. DoS > 0.9 indicates near-capacity operation per SCATS design standards.
DoS = v/c (volume / saturation flow)
d = C(1−g)² / [2(1−g·DoS)] + d₂
Fusion Bottleneck Score
Composite 0–100 score combines TomTom link congestion with SCATS intersection saturation. Spillback is flagged when a link's downstream intersection has DoS > 0.9 AND link TTI > 1.4 — confirming the queue has extended upstream past the stop line.
BS = 50×min((TTI−1)/1.0, 1) + 50×min(DoS, 1)
Data sources: TomTom probe vehicles (GPS traces, hourly aggregation) · SCATS detector counts (cycle-by-cycle, hourly DoS) · LOS thresholds from HCM 6th Edition · Demonstration dataset: typical Auckland CBD weekday pattern