Overview
Most congested approaches in range (by stop rate)
| Approach | Junction | Stop rate | Avg stops | Travel time | Probe Count |
|---|
Intersection performance scorecard
Intersections / approaches / movements graded and ranked for the selected window to prioritize signal retiming. The grade is built from directly-measured signal-performance metrics: control delay, AOG, split failure, and stops.
Parameter measurements
Per-approach metrics for the selected range. Click a row for its 24-hour profile.
| Approach | Junc | Probe Count | Traj/day | Stop % | AOG % | Avg stops | Split % | Travel s | Delay s | Speed | Decel | Accel | L% | T% | R% |
|---|
24-hour profile
Time-series trends
Daily trend, hour-of-day and weekday patterns for the selected range.
Daily trend
Hour-of-day profile
By weekday (weekday vs weekend)
Monthly comparison
Overlay each month's hour-of-day profile for the selected route/approach and metric above — tick months to compare them on one chart. Respects the weekday filter.
Period comparison & time-slot analysis
Per-weekday evaluation
Every parameter averaged for each day of the week, over the selected date range. Worst day per metric is shown in red. (This view ignores the weekday/weekend toggle.)
| Day | Days | Probe Count | Probes/day | Avg stops | Stop % | Travel s | Delay s | Speed | AOG % | Split % | L% | T% | R% |
|---|
Anomaly detection
Anomalies per day in range
Flagged events
| Date | Hour | Approach | Junc | Metric | Value | Baseline | z | Dir | Severity | Why |
|---|
Multi-dimensional anomaly analysis
Classified events
| Date | Hour | Approach | Move | Junc | Corridor | Vol z | Delay z | Stops z | Speed z | Diagnosis | Severity |
|---|
Corridor coordination (green-wave)
How this works — methodology
What it measures. Coordination (green-wave) quality via through-movement AOG — Arrival on Green, the share of through vehicles that arrive on green and don't stop: (through vehicles − through vehicles that stopped) ÷ through vehicles × 100. High AOG = cars flow through = well-coordinated; low AOG = everyone stops. Only the through movement is used, since coordination is about the main corridor flow.
Travel order. For the chosen corridor + direction, the through approach at each intersection is sorted by longitude so it reads in the order a driver passes them (west→east, reversed for westbound).
The four panels. Strip = each intersection's AOG in travel order (red outline = weak link, the lowest-AOG intersection); corridor average is volume-weighted. Hour heatmap = AOG split by hour × intersection — a vertical red band shows when the wave breaks, a red row shows which intersection is chronically bad. Map = intersections at their real coordinates coloured by AOG, connected in travel order. Historical = corridor AOG per day.
Abnormal-day flagging. Each day's corridor AOG is compared to the same-weekday baseline in the window (mean and standard deviation of that weekday's AOG). A day is flagged when its z-score = (day − weekday mean) ÷ std dev is ≤ −2 (AOG at least 2σ below normal — downside only, since we care about coordination breaking). Guards: a day needs ≥30 through vehicles to count, and a weekday group needs ≥4 comparable days to be judged.
Fields used. From the movement dataset: through vehicle count and through "has-stops" count per intersection/hour/day, plus each approach's lat/lon for the map. Nothing else feeds this tab.
Corridor map — AOG by intersection
Markers coloured by through AOG (green good → red poor), line follows travel order. The reddest marker is the coordination weak link.
AOG analysis — problematic hours & abnormal coordination
Abnormal AOG events
Each row = one intersection-hour-day whose through AOG fell below its own same-weekday+hour norm. Click any header to sort. These unusual failures often fall in lower-volume periods (nights/weekends), distinct from the structural peak-hour problems in the charts above.
| Date | Day | Hour | Intersection | Junc | Corridor | AOG | Baseline | z | Thru N |
|---|