ETA accuracy benchmark · 🇪🇸 Valencia

How Port Flow's predicted ETA compares to each vessel's own broadcast ETA at Valencia, measured on closed voyages over the last 90 days. We publish the number — good or bad. Lower error is better.

Port Flow predicted ETA (RMSE)

20.82 h

MAE 25.44 h · 877 voyages

Broadcast ETA (RMSE)

63.35 h

crew-declared, 556 voyages (same set)

Our advantage

67.1 %

lower error than the broadcast ETA

Broadcast ETAs more than 7 days from the actual arrival are excluded as AIS sentinel / placeholder values (315 excluded). Both figures are computed on the same remaining voyage set — apples to apples.

Closed voyages

error = predicted/broadcast minus actual arrival (hours)


255806064	general-cargo	08/26, 09:56 PM UTC	-0.05 h	-555.95 h
357444000	container	08/26, 08:12 PM UTC	-0.30 h	-30.20 h
354744000	container	08/26, 05:44 PM UTC	-0.15 h	1.27 h
352965000	container	08/26, 02:52 PM UTC	-1.42 h	-770.87 h
538005959	container	08/26, 02:45 PM UTC	-0.05 h	-1339.76 h
244985000	container	08/26, 12:10 PM UTC	-0.62 h	-479.18 h
440981000	container	08/26, 10:50 AM UTC	0.09 h	-2.84 h
305741000	container	08/26, 10:05 AM UTC	0.25 h	-39.10 h
256744000	container	08/26, 07:18 AM UTC	0.49 h	-3.31 h
249783000	container	08/26, 03:41 AM UTC	0.38 h	-1940.52 h
352002508	product	08/26, 02:39 AM UTC	-0.27 h	0.84 h
249850000	container	08/26, 01:47 AM UTC	0.21 h	-1196.79 h
636015665	container	08/25, 08:00 PM UTC	-0.80 h	-982.00 h
249930000	product	08/25, 06:12 PM UTC	-0.34 h	-79.70 h
277550000	container	08/25, 02:33 PM UTC	-0.30 h	-742.06 h
636021362	container	08/25, 12:38 PM UTC	-2.39 h	-1230.64 h
249930000	product	08/25, 12:15 PM UTC	0.15 h	-73.76 h
428002000	container	08/25, 11:23 AM UTC	0.15 h	-759.39 h
247383300	general-cargo	08/25, 10:52 AM UTC	0.26 h	-397.88 h
247384100	general-cargo	08/25, 05:43 AM UTC	0.17 h	-570.23 h
235067987	general-cargo	08/25, 05:40 AM UTC	-0.27 h	-2.67 h
249813000	container	08/25, 12:06 AM UTC	0.02 h	-1839.11 h
271001058	general-cargo	08/24, 05:41 PM UTC	0.71 h	-756.20 h
248302000	general-cargo	08/24, 02:06 PM UTC	-0.69 h	-622.11 h
212641000	container	08/24, 01:18 PM UTC	0.27 h	—
563024600	container	08/24, 09:16 AM UTC	0.29 h	-738.78 h
256328000	product	08/24, 08:41 AM UTC	-0.62 h	-53.69 h
255916056	container	08/24, 06:33 AM UTC	0.07 h	-1838.56 h
224127820	product	08/24, 05:58 AM UTC	0.36 h	3346.02 h
477967800	general-cargo	08/24, 04:48 AM UTC	0.57 h	-1.81 h
636018712	general-cargo	08/24, 03:05 AM UTC	-0.11 h	-1659.60 h
229625000	container	08/24, 12:14 AM UTC	-1.57 h	-3.24 h
636021596	container	08/23, 10:58 PM UTC	0.72 h	-1289.97 h
255993000	container	08/23, 07:33 PM UTC	0.34 h	-8.56 h
636021337	container	08/23, 07:11 PM UTC	-0.63 h	-2129.19 h
212641000	container	08/23, 06:44 PM UTC	-0.40 h	-590.75 h
255993000	container	08/23, 01:30 PM UTC	-0.30 h	-2.51 h
314851000	container	08/23, 09:01 AM UTC	-0.47 h	-109.03 h
249930000	product	08/23, 08:56 AM UTC	0.03 h	-22.44 h
224127820	product	08/23, 05:09 AM UTC	-11.75 h	3370.84 h
247039400	container	08/23, 04:25 AM UTC	-0.05 h	-0.42 h
212136000	container	08/23, 03:12 AM UTC	-0.16 h	-24.21 h
249930000	product	08/23, 01:06 AM UTC	0.41 h	-14.60 h
636020326	general-cargo	08/23, 12:31 AM UTC	0.07 h	-1607.02 h
247383600	general-cargo	08/22, 11:41 PM UTC	0.47 h	-91.19 h
538007610	container	08/22, 10:28 PM UTC	0.46 h	-722.47 h
255805984	container	08/22, 08:09 PM UTC	0.47 h	-554.16 h
255805844	container	08/22, 05:59 PM UTC	0.38 h	-583.99 h
636092930	general-cargo	08/22, 03:49 PM UTC	-0.21 h	-1020.82 h
352898711	container	08/22, 02:52 PM UTC	-93.36 h	-1088.88 h

See it on a live desk

Predicted ETAs, congestion and sanctions screening across 51 ports — try it in one click, no signup.

Open dashboard See pricing →

How it's measured

Predicted ETA: distance-to-berth over speed-made-good, refreshed on every AIS position, with a seasonal correction from rolling medians.
Broadcast ETA: the crew-entered ETA in the vessel's AIS message — the de-facto industry default.
Both are compared to the actual arrival timestamp on the same closed voyages. We report RMSE (penalises large misses) and MAE (typical error).
Broadcast ETAs more than 7 days off are dropped as AIS sentinel / placeholder values, and our model is scored on that same set — so the comparison is honest, not cherry-picked.
Coverage is public AIS only; ports with weak coverage show fewer closed voyages. We'd rather show a small honest sample than a big misleading one.