Skip to content

CONCEPT Cited by 1 source

Estimated Time of Arrival (ETA)

Estimated Time of Arrival (ETA) is the predicted clock time at which a trip (vehicle, delivery, transit, courier) will reach its destination, given the origin, the route, live road conditions, and trip-specific features. In consumer mapping products it is the load-bearing number the rest of the UX hangs off — route ranking, turn-by-turn recalibration, sharing "I'll be there at 6:42".

Stub page. The wiki has one ingested source touching ETA so far — Google's HOV-specific ETAs post, which only surfaces the intro. This page will expand substantially as Maps / Uber / DoorDash / logistics ETA posts are ingested; the framing below is a scaffold, not a survey.

Why ETAs are hard

  • Route-conditioned, not just origin-destination. The same pair of endpoints has many legal routes; each has a different ETA.
  • Live traffic is a moving target. Historical priors (time of day, day of week, weather) must be fused with live probe data (vehicles reporting speed / position) under seconds-of-latency freshness requirements.
  • Lane type and trip category matter. A single ETA conditioned only on route + time mispredicts when the road network has heterogeneous lanes available to some trips but not others. Google's 2025-06-30 HOV post is the canonical wiki instance of this failure mode: HOV lanes in Utah's Salt Lake Valley averaged 68.18 mph vs. 58.60 mph in general lanes (~16% faster) — large enough that folding HOV and non-HOV traffic into one ETA materially biases both (Source: sources/2025-06-30-google-hov-specific-etas-google-maps).
  • Serving latency budget is tight. ETAs are computed inline in the navigation UI and re-computed continuously as the driver moves.

Design moves seen so far

  • Split ETA surfaces by trip category. Rather than adding category features to a single model and hoping it learns the split, Google Maps HOV ETAs use a dedicated ETA surface conditioned on a trip classifier output (HOV vs. non-HOV). The classifier is the enabling artifact that the specialised surface depends on (Source: sources/2025-06-30-google-hov-specific-etas-google-maps).
  • Expose the split in the UI. Route options (e.g., "include carpool lanes") let users opt into the category, so the ETA surface being selected is transparent and controllable.

Seen in

Last updated · 200 distilled / 1,178 read