The numbers behind pricing a road — where the passengers are, what a route is worth, which vehicles are deviating, and what the fleet's claim rate is running at. Every figure on this page comes from the live network; nothing here is illustrative.
Observed passenger counts at real ranks. This is the demand side of the network — where the volume physically is, before anyone models anything.
The origin–destination graph, weighted by passengers. The hubs below are the corridors that carry the network — the ones worth instrumenting first.
Revenue proxy against measured risk exposure, per route. This is the sentence "we can price this road" turned into a number.
The fleet's incident rate per kilometre, with the warm-up guard that stops a freshly restarted fleet reporting a nonsense rate off a handful of kilometres.
Vehicles deviating from their own established band, or off the known network entirely. Watching is not flagging — a vehicle only appears once it is genuinely out.
Learned activity density by zone and hour. It is advisory: it positions supply, it does not dispatch anyone.
How to read this. These are measurements, not projections. Rank demand is observed survey volume; route economics is a revenue proxy against measured exposure, not billed revenue; the risk forecast is a rate per kilometre and says so when it has not yet seen enough distance to be meaningful. Where a number is not yet trustworthy this page says so rather than rounding it into confidence.
© OpenStreetMap