Skip to content

Auckland Transport data via eolas

Auckland Transport (AT) operates Auckland's public transport network — bus, ferry, train, plus the cycle network and park-and-ride facilities. eolas serves 10 datasets from AT — the spatial layers behind every "how do I get from A to B" query in Auckland.

If you're doing transport-planning, mode-share analysis, accessibility studies, or building a transit-aware Auckland app — this is the source.

For Auckland's planning data (district plan, hazards, heritage), see Auckland Council. AT is a separate Council-Controlled Organisation (CCO) with its own data portal.

What's in the catalogue

Public transport

Dataset Description
akt_bus_route All AT bus routes — polyline geometry + route number, direction.
akt_bus_stop All AT bus stops — point geometry + accessibility flags, shelter info.
akt_school_bus_route School bus routes (separate from regular network).
akt_school_bus_stop School bus stops.
akt_ferry_route Ferry routes — polyline + operator, service frequency.
akt_ferry_stop Ferry terminals + landings — point.
akt_train_station Train stations — point geometry + platforms, parking, accessibility.
akt_park_and_ride Park-and-ride facilities — point + capacity, mode connections.

Active modes

Dataset Description
akt_cycle_facility_network All Auckland cycle network infrastructure — separated paths, shared paths, on-road lanes, sharrows.
akt_bridges Bridges in the AT-managed road network.

Refresh schedule

Weekly, Wednesday morning NZ time. AT publishes via the AT Open Data Portal (ArcGIS).

Bus routes and stops can change quarterly with service reviews; the cycle network grows as new infrastructure is built (typically several updates per year). Train stations are stable; ferry routes change rarely.

For real-time transit data (vehicle locations, arrival predictions), use AT's GTFS-RT feed — not in eolas.

meta = client.info("akt_bus_route")
meta["last_refreshed_at"]
meta["source_last_modified_at"]

License

All AT data is published under CC-BY 4.0. Commercial use is fine; attribution required.

Recommended attribution: "Source: Auckland Transport, served via eolas (eolas.fyi). CC-BY 4.0."

Common patterns

Plot the entire Auckland network

import matplotlib.pyplot as plt

buses = client.auckland_transport("akt_bus_route", as_sf=True)
trains = client.auckland_transport("akt_train_station", as_sf=True)
ferries = client.auckland_transport("akt_ferry_route", as_sf=True)
cycle = client.auckland_transport("akt_cycle_facility_network", as_sf=True)

fig, ax = plt.subplots(figsize=(10, 12))
buses.plot(ax=ax, color="lightblue", linewidth=0.4, label="Bus")
cycle.plot(ax=ax, color="green", linewidth=0.6, label="Cycle")
ferries.plot(ax=ax, color="navy", linewidth=1.0, label="Ferry")
trains.plot(ax=ax, color="red", markersize=20, label="Train station")
ax.set_title("Auckland public transport network")
ax.legend()
plt.show()

library(sf)
library(ggplot2)

buses <- eolas_get_auckland_transport("akt_bus_route", as_sf = TRUE)
trains <- eolas_get_auckland_transport("akt_train_station", as_sf = TRUE)
cycle <- eolas_get_auckland_transport("akt_cycle_facility_network", as_sf = TRUE)

ggplot() +
  geom_sf(data = buses, linewidth = 0.2, colour = "lightblue") +
  geom_sf(data = cycle, linewidth = 0.5, colour = "green") +
  geom_sf(data = trains, size = 1.5, colour = "red") +
  theme_void() +
  labs(title = "Auckland PT network")

Bus-stop accessibility for a property

import geopandas as gpd
from shapely.geometry import Point

# Walking accessibility = bus stops within 400m of a point
pt = gpd.GeoSeries([Point(174.7633, -36.8485)], crs="EPSG:4326").to_crs("EPSG:2193")
stops = client.auckland_transport("akt_bus_stop", as_sf=True).to_crs("EPSG:2193")

nearby = stops[stops.geometry.distance(pt.iloc[0]) <= 400]
print(f"Bus stops within 400m: {len(nearby)}")

Cycle-network density by suburb

import geopandas as gpd

cycle = client.auckland_transport("akt_cycle_facility_network", as_sf=True)
suburbs = client.linz("nz_suburbs_and_localities", as_sf=True)
# Cycle km per suburb
cycle = cycle.to_crs("EPSG:2193")
cycle["length_m"] = cycle.geometry.length
suburb_with_cycle = gpd.sjoin(cycle, suburbs, how="inner", predicate="intersects")
top = suburb_with_cycle.groupby("name")["length_m"].sum().sort_values(ascending=False).head(15)
print(top.div(1000).round(1).astype(str) + " km")

Source-specific notes

  • AT is a CCO, not the Council: data ownership + publication is separate from Auckland Council. AT also runs the road network (AT-managed roads), shoulder of NZTA state highways within Auckland.
  • Routes vs services vs trips: akt_bus_route is the spatial route a service follows. Frequency, timetable, and live arrivals are in the GTFS feed (not in eolas) — they reference a route_id that matches AT's GTFS data.
  • Cycle-facility types: the facility_type field distinguishes separated paths (most common in recent infrastructure), shared paths, on-road lanes, and sharrows (shared with cars). Don't treat them as equivalent for accessibility analysis.
  • Stops + stations have accessibility flags: wheelchair_accessible, shelter, bike_parking, etc. Useful for equity-of-access analysis.
  • Not in eolas: GTFS schedule data, GTFS-RT (real-time), HOP card usage stats. AT publishes GTFS at dev-portal.at.govt.nz — better suited as a streaming dataset than a batch table.
  • Train stations are operated by AT but the rail network is KiwiRail's: AT owns Britomart + most platforms; the lines/signals are KiwiRail. For the rail line geometry (where the tracks actually run), use LINZ rail centrelines.

Where to find more