ETA accuracy benchmark · 🇺🇸 Los Angeles

How Port Flow's predicted ETA compares to each vessel's own broadcast ETA at Los Angeles, measured on closed voyages over the last 7 days. We publish the number — good or bad. Lower error is better.

Port Flow predicted ETA (RMSE)

16.79 h

MAE 166.36 h · 76 voyages

Broadcast ETA (RMSE)

29.10 h

crew-declared, 32 voyages (same set)

Our advantage

42.3 %

lower error than the broadcast ETA

Broadcast ETAs more than 7 days from the actual arrival are excluded as AIS sentinel / placeholder values (44 excluded). Both figures are computed on the same remaining voyage set — apples to apples.

Closed voyages

error = predicted/broadcast minus actual arrival (hours)


538004202	container	08/26, 11:07 PM UTC	0.73 h	-1.12 h
357795000	container	08/26, 10:45 PM UTC	1.28 h	-1946.26 h
636023639	container	08/26, 10:00 PM UTC	0.58 h	3.00 h
636020671	container	08/26, 09:02 PM UTC	2.25 h	-24.05 h
566448000	container	08/26, 08:44 PM UTC	1.08 h	-20.25 h
636091487	container	08/26, 05:49 PM UTC	-0.07 h	-426.83 h
352006501	product	08/26, 05:04 PM UTC	0.27 h	-1.07 h
636020846	container	08/26, 02:48 PM UTC	0.88 h	-1.31 h
256071000	container	08/26, 02:14 PM UTC	-0.29 h	-890.25 h
566448000	container	08/26, 02:09 PM UTC	2.02 h	-13.66 h
636024513	container	08/26, 01:35 PM UTC	0.57 h	-1.09 h
367719640	container	08/26, 08:34 AM UTC	-0.28 h	-5.58 h
538011148	product	08/26, 12:28 AM UTC	0.88 h	-73.47 h
538011148	product	08/25, 11:13 PM UTC	-54.59 h	-1367.21 h
354120000	container	08/25, 10:59 PM UTC	-0.43 h	-912.99 h
565807000	container	08/25, 09:52 PM UTC	0.40 h	-0.88 h
636020671	container	08/25, 09:47 PM UTC	0.64 h	-0.79 h
538011714	container	08/25, 09:12 PM UTC	0.54 h	-510.21 h
477056400	general-cargo	08/25, 09:01 PM UTC	1.38 h	-1050.03 h
563276200	product	08/25, 05:17 PM UTC	-0.12 h	-6.29 h
311056900	container	08/25, 03:55 PM UTC	1.52 h	-514.92 h
354004000	container	08/25, 02:43 PM UTC	1.08 h	-985.73 h
232018901	container	08/25, 01:37 PM UTC	-0.64 h	-1.62 h
636024272	container	08/25, 01:28 PM UTC	0.06 h	-44.47 h
367641230	container	08/25, 01:12 PM UTC	0.27 h	-537.21 h
369006000	container	08/25, 12:49 PM UTC	-0.01 h	-1195.83 h
367719640	container	08/25, 08:22 AM UTC	-0.16 h	-5.87 h
366904000	product	08/25, 02:27 AM UTC	3.16 h	-481.96 h
338110000	product	08/25, 01:45 AM UTC	1.44 h	-34.76 h
219314000	container	08/25, 12:24 AM UTC	0.70 h	-1.40 h
565402000	container	08/25, 12:17 AM UTC	0.88 h	-634.29 h
636024494	container	08/24, 11:55 PM UTC	-0.53 h	13.07 h
565819000	container	08/24, 09:57 PM UTC	0.36 h	-849.95 h
219265000	product	08/24, 09:38 PM UTC	1.38 h	-75.65 h
319130700	product	08/24, 01:42 PM UTC	-0.28 h	-1898.70 h
357147000	container	08/24, 01:39 PM UTC	1.42 h	-1.16 h
311695000	container	08/24, 01:11 PM UTC	0.53 h	-1.20 h
368359000	container	08/24, 11:57 AM UTC	-0.71 h	-595.95 h
372628000	product	08/24, 10:16 AM UTC	-0.62 h	-1470.18 h
636024119	container	08/24, 09:34 AM UTC	-0.29 h	-1333.57 h
352004278	product	08/24, 05:00 AM UTC	2.44 h	-367.01 h
441502000	general-cargo	08/24, 01:10 AM UTC	0.10 h	-671.18 h
366948190	product	08/23, 11:53 PM UTC	1.69 h	-2122.89 h
440189000	general-cargo	08/23, 10:53 PM UTC	1.04 h	-1137.89 h
636023071	container	08/23, 09:10 PM UTC	0.56 h	-1.17 h
369207000	container	08/23, 08:43 PM UTC	1.45 h	-666.73 h
538007701	container	08/23, 07:45 PM UTC	2.08 h	-19.74 h
338110000	product	08/23, 06:23 PM UTC	1.37 h	-785.89 h
538012038	product	08/23, 02:56 PM UTC	0.41 h	-0.94 h
368237190	container	08/23, 02:27 PM UTC	-0.27 h	-38.13 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.