ETA accuracy benchmark · 🇮🇹 Trieste

How Port Flow's predicted ETA compares to each vessel's own broadcast ETA at Trieste, 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)

22.28 h

MAE 26.58 h · 731 voyages

Broadcast ETA (RMSE)

64.23 h

crew-declared, 493 voyages (same set)

Our advantage

65.3 %

lower error than the broadcast ETA

Broadcast ETAs more than 7 days from the actual arrival are excluded as AIS sentinel / placeholder values (208 excluded). Both figures are computed on the same remaining voyage set — apples to apples.

Closed voyages

error = predicted/broadcast minus actual arrival (hours)


636015908	container	08/26, 11:26 PM UTC	2.55 h	-2.45 h
238255000	product	08/26, 07:03 PM UTC	1.98 h	-7.30 h
376304000	container	08/26, 06:35 PM UTC	3.06 h	-2.60 h
253000119	container	08/26, 03:58 PM UTC	1.11 h	-653.97 h
311000765	product	08/26, 02:33 PM UTC	-1.08 h	-505.55 h
636022914	container	08/26, 01:59 PM UTC	2.64 h	-1275.49 h
247383900	general-cargo	08/26, 10:36 AM UTC	0.96 h	-0.62 h
271049958	general-cargo	08/26, 10:08 AM UTC	-1.59 h	-21.64 h
271000793	container	08/26, 07:39 AM UTC	-0.96 h	-492.67 h
271043165	container	08/26, 05:48 AM UTC	0.01 h	-0.56 h
636021665	container	08/26, 03:16 AM UTC	-5.38 h	-6.27 h
636018324	container	08/25, 10:31 PM UTC	2.90 h	-586.52 h
538002852	product	08/25, 09:48 PM UTC	2.46 h	-1.31 h
538005389	product	08/25, 08:33 PM UTC	1.51 h	-16.56 h
354149000	container	08/25, 08:02 PM UTC	2.53 h	-1.04 h
247383800	container	08/25, 03:59 PM UTC	-9.91 h	-203.00 h
271049958	general-cargo	08/25, 12:34 PM UTC	0.78 h	-602.58 h
248039000	container	08/25, 12:29 PM UTC	2.49 h	-2091.50 h
636021506	container	08/25, 05:54 AM UTC	2.27 h	-2.91 h
271000747	container	08/25, 05:16 AM UTC	0.26 h	—
271046601	general-cargo	08/25, 05:09 AM UTC	-0.84 h	-86.91 h
538005826	product	08/25, 12:18 AM UTC	0.66 h	-31.30 h
636024846	container	08/24, 08:57 PM UTC	3.59 h	-16.96 h
247418500	container	08/24, 04:50 PM UTC	-0.00 h	-0.33 h
271043164	general-cargo	08/24, 03:55 PM UTC	0.26 h	-0.93 h
255806128	container	08/24, 01:40 PM UTC	0.01 h	-1.68 h
477708600	container	08/24, 01:18 PM UTC	2.52 h	-8.31 h
677021400	container	08/24, 11:54 AM UTC	3.73 h	-785.91 h
563221900	container	08/24, 10:39 AM UTC	-0.49 h	-1.65 h
248108000	product	08/24, 08:51 AM UTC	-0.47 h	-17.85 h
242494400	container	08/24, 08:16 AM UTC	2.43 h	-1630.27 h
245261000	container	08/24, 05:13 AM UTC	-0.47 h	-1.23 h
271000623	general-cargo	08/24, 04:50 AM UTC	-0.87 h	-0.59 h
636024014	container	08/24, 04:46 AM UTC	-0.40 h	-13.03 h
305739000	container	08/23, 06:47 PM UTC	3.58 h	-1987.80 h
563221900	container	08/23, 04:05 PM UTC	1.15 h	-595.09 h
271000792	general-cargo	08/23, 05:39 AM UTC	-0.20 h	-522.66 h
477010100	product	08/23, 05:13 AM UTC	0.70 h	-7.22 h
636024014	container	08/23, 02:02 AM UTC	2.97 h	-640.04 h
215068000	product	08/22, 08:56 PM UTC	0.67 h	-1812.95 h
538005400	product	08/22, 02:56 PM UTC	-1673.21 h	-1624.94 h
247453500	general-cargo	08/22, 02:45 PM UTC	-542.88 h	-454.76 h
211882930	general-cargo	08/22, 02:42 PM UTC	-2572.83 h	-1617.71 h
538012851	container	08/22, 02:41 PM UTC	-450.58 h	-452.19 h
636014069	product	08/22, 02:41 PM UTC	-863.62 h	-785.69 h
538009721	container	08/22, 02:40 PM UTC	-728.79 h	-634.68 h
256010000	container	08/22, 02:39 PM UTC	-1028.32 h	-957.67 h
255916299	container	08/19, 05:17 AM UTC	0.54 h	-307.30 h
247383900	container	08/19, 05:00 AM UTC	0.84 h	-430.50 h
249787000	container	08/19, 04:34 AM UTC	1.74 h	-1952.57 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.