Pixl Dispatch — Curb Coverage Sim

Loading real data…

transit busridesharePixl route

fresh

overdue

Transit buses across the day

shaded = paid parking hours · drag to scrub

Transit buses as mobile sensors. Every Winnipeg Transit bus carries two curbside cameras, so as it runs its real route on the real weekday GTFS schedule it passively scans the curb segments it passes (both sides). Fresh via transit is the share of the currently-fresh curb that a bus scanned most recently — it climbs while buses are running and falls to zero overnight (the day timeline shows the live bus count, peaking at morning and afternoon rush). About 40% of these curb segments sit on a bus route, so transit keeps that share fresh almost for free during service hours; Pixl scan routes cover the rest. Toggle the transit sensor layer off (or drop routes to zero) to see how much each source carries. Paid-parking hours (shaded, ~08:00–17:30 assumed) are the priority, but the full 24h is tracked for planning and wayfinding.

Opportunistic rideshare. Uber/Lyft drivers running the Pixl app for a small fee scan the curb as they drive; every vehicle here captures both curbs of the street it's on (single-side capture only matters for licence-plate reading — an ALPR concern — whereas occupancy reads fine from either side). Fresh via rideshare is their share of the fresh curb. Rides are modelled from real geography: intra-downtown hops, trips in and out through the cordon, and a heavy airport↔hotel corridor, with demand weighted to evenings and late night — exactly when transit thins out, so the two sources are complementary. Slide Rideshare cars to model how many drivers opt in. Curb data = City of Winnipeg WPA paid segments; streets = OpenStreetMap; buses = Winnipeg Transit GTFS; rideshare rides are a synthetic model on the real network. Pixl scan routes are short batched sweeps (~5–10 min) bundling several overdue blocks, offered to the rideshare pool to run between fares — so the number of active routes is demand-driven (the slider is a budget ceiling, not a headcount), and it self-throttles when buses and rideshare already keep the curb fresh. Route-building is the greedy MVP; production swaps in OR-Tools arc-routing over an OSRM matrix. The Pixl driver cost card totals the kilometres the assigned route drivers cover in a simulated day (buses and rideshare aren't paid per km) × the pay-rate slider — a first-order estimate of daily data-collection cost.