Bus comparisons
The same stretch. Matched bus, road and timetable comparisons.
Bus 190 and TomTom car-route estimates between Hogarth Roundabout and Stamford Brook Bus Garage. Mon–Fri pooled hourly medians, excluding England and Wales bank holidays.
Same bus corridor · separate from the Road traffic tab
Here, both bus and car times cover the same bus-stop-to-bus-stop corridor between Hogarth Roundabout and Stamford Brook Bus Garage. The Road traffic tab monitors different, wider Chiswick Lane routes: its journey times are not used in these paired bars.
We began collecting separate TomTom car-route estimates for this bus corridor on 7 November 2026, alongside bus tracking. Each qualifying bus start triggers a live TomTom request, saved at the time with traffic-aware and no-traffic reference times. These car figures are modelled estimates, not inferred from the wider road routes or reconstructed for earlier dates. The paired bus-versus-road bars include only completed bus journeys with a valid saved TomTom match; earlier or missing comparisons remain absent.
Loading matched journeys…
Mon–Fri · bus and road side by side
Both bars use exactly the same matched journeys in each hour. Each value is a median of individual journey values, not a median of weekday medians. Shared vertical scale across directions; empty hours stay empty. Swipe charts sideways on mobile. Tap or focus bars for details.
Bus 190 · captured vs scheduled duration
The same stop-to-stop corridor, compared with each matched bus journey’s archived TfL timetable. The date range and estimates checkbox above apply to all charts below; the Ratio / Time / Speed buttons apply only to the bus-versus-road charts.
Loading timetable matches…
How much longer than scheduled? · 07:00–19:00
Starts from 07:00 up to but not including 19:00 London time, across all selected days. Each direction totals 100%, using half-minute bins. Positive means a longer duration than scheduled; negative means shorter. Colours show extra duration: green below 0 minutes; orange from 0 to 1 minute; red over 1 up to 3 minutes; dark red over 3 minutes. Summary cards use the same filtered histogram sample. Worst 10% is the average extra duration among the longest differences, rounding the group size up to a whole journey; it is not a percentile cut-off. Longest extra time is the maximum captured-minus-scheduled duration in that sample.
Mon–Fri pooled · scheduled reference
Bars: median captured minutes. White dotted segments: median scheduled minutes for exactly the same matched journeys in each hour. Labels show the signed difference of these medians in minutes, not sample counts or the median of individual differences. Weekends and England/Wales bank holidays excluded; all hours retained. Mobile starts at 06:00–20:00; swipe for other hours. Colours use the website’s congestion ratio: ≤120% green, >120–150% yellow, >150% red to dark red; grey means no valid reference ratio.
Individual journeys · scheduled vs captured
Each dot is one matched journey across all selected days/hours. Blue: northbound; yellow: southbound. White dotted line: equal duration. Points above the line took longer than scheduled. Tap or focus for details.
Sources, matching and limitations
Schedules come from TfL’s public iBus static archive, not historical measured performance. Captured trip identifiers are matched privately to timetable journey identifiers, direction and valid operating date. Repeated daily identifiers are resolved using the closest scheduled corridor start, including after-midnight services; ambiguous matches or start differences above two hours are withheld. No vehicle or trip identifiers are published. Unmatched captures stay excluded, never given a guessed schedule. New or changed timetable identifiers may not yet be available in the saved archive.
Both durations are unadjusted arrival-to-arrival times between the same directional stops: no one-minute subtraction. Only the bar colours use the separately adjusted congestion ratio described below. Captured times are inferred from TfL predictions, not confirmed arrivals. Extended-window estimates use an endpoint prediction made over 90 and up to 150 seconds ahead, rather than within 90 seconds; they remain included unless you untick the shared checkbox.
Extra duration is not lateness at a stop. These descriptive comparisons do not identify the cause of delays or establish the effect of the Weltje Road restriction. Small or incomplete samples can be unstable; today is partial.