ETA accuracy benchmark · 🇺🇸 New Orleans / Lower Mississippi
How Port Flow's predicted ETA compares to each vessel's own broadcast ETA at New Orleans / Lower Mississippi, 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)
205.46 h
MAE 59.94 h · 56 voyages
Broadcast ETA (RMSE)
Building…
only 15 comparable voyages so far (need 20)
Gap
—
benchmark builds as more voyages close
Closed voyages
error = predicted/broadcast minus actual arrival (hours)| 367006230 | product | 08/24, 01:13 PM UTC | 1.90 h | — |
| 367379810 | container | 08/24, 01:06 PM UTC | -3.20 h | — |
| 636020525 | container | 08/24, 12:56 PM UTC | 1.14 h | -11.95 h |
| 367006230 | product | 08/24, 12:28 PM UTC | -0.63 h | — |
| 367379810 | container | 08/24, 11:24 AM UTC | -1.19 h | — |
| 636016986 | container | 08/24, 10:16 AM UTC | -4.49 h | -11.27 h |
| 367379810 | container | 08/24, 08:57 AM UTC | 3.17 h | — |
| 538011265 | container | 08/24, 08:51 AM UTC | 12.97 h | -44.85 h |
| 368253470 | container | 08/24, 06:28 AM UTC | -20.57 h | -20.48 h |
| 563152900 | product | 08/24, 04:43 AM UTC | 7.82 h | — |
| 636021440 | general-cargo | 08/23, 03:06 PM UTC | -4.17 h | -9.11 h |
| 367379810 | container | 08/23, 12:28 PM UTC | 8.69 h | — |
| 367379810 | container | 08/23, 09:56 AM UTC | 7.78 h | — |
| 367379810 | container | 08/23, 09:50 AM UTC | 4.18 h | — |
| 636021279 | container | 08/23, 06:43 AM UTC | 14.43 h | — |
| 367006230 | product | 08/23, 05:44 AM UTC | -530.50 h | — |
| 538012307 | container | 08/23, 05:02 AM UTC | 0.59 h | — |
| 563286200 | container | 08/04, 06:05 AM UTC | -1298.09 h | -1433.60 h |
| 563331700 | container | 08/04, 05:52 AM UTC | -140.26 h | — |
| 374106000 | container | 08/01, 02:21 AM UTC | -61.38 h | -69.35 h |
| 538004246 | product | 07/31, 12:47 PM UTC | -4.28 h | — |
| 368253470 | container | 07/31, 12:29 PM UTC | -0.81 h | — |
| 367379810 | container | 07/31, 09:13 AM UTC | -2.68 h | — |
| 636016185 | general-cargo | 07/31, 09:03 AM UTC | -4.51 h | -11.06 h |
| 367379810 | container | 07/31, 07:25 AM UTC | 8.94 h | — |
| 538011014 | product | 07/31, 06:49 AM UTC | 26.36 h | -16.82 h |
| 367379810 | container | 07/31, 06:31 AM UTC | 28.95 h | — |
| 538008079 | product | 07/31, 05:40 AM UTC | 21.15 h | — |
| 367379810 | container | 07/31, 04:43 AM UTC | 22.06 h | — |
| 636021119 | product | 07/31, 04:16 AM UTC | -40.24 h | — |
| 367379810 | container | 07/31, 04:04 AM UTC | 3.72 h | — |
| 255726000 | container | 07/31, 03:41 AM UTC | -135.33 h | — |
| 636017489 | container | 07/31, 03:35 AM UTC | -59.24 h | -72.60 h |
| 563238600 | product | 07/31, 03:22 AM UTC | -63.88 h | — |
| 636021992 | container | 07/31, 03:10 AM UTC | -17.97 h | — |
| 354455000 | container | 07/29, 10:04 AM UTC | 3.16 h | — |
| 367379810 | container | 07/29, 09:21 AM UTC | 4.49 h | — |
| 566637000 | container | 07/29, 09:02 AM UTC | -17.29 h | — |
| 636019582 | container | 07/29, 08:51 AM UTC | 3.94 h | — |
| 374597000 | product | 07/29, 08:44 AM UTC | -1.94 h | — |
| 538011987 | product | 07/28, 12:29 PM UTC | 8.48 h | -10.49 h |
| 367006230 | product | 07/28, 12:24 PM UTC | 0.39 h | — |
| 367379810 | container | 07/28, 09:46 AM UTC | 5.66 h | — |
| 636023343 | container | 07/28, 09:15 AM UTC | 3.65 h | -10.26 h |
| 355216000 | container | 07/19, 09:15 AM UTC | 3.12 h | -10.25 h |
| 372697000 | container | 07/19, 05:49 AM UTC | -579.03 h | — |
| 636013521 | container | 06/26, 05:48 AM UTC | -16.32 h | -762.82 h |
| 538012009 | product | 06/25, 07:03 AM UTC | -1.99 h | — |
| 259099000 | container | 06/25, 04:54 AM UTC | -2.83 h | — |
| 636021440 | general-cargo | 06/21, 04:12 AM UTC | -4.91 h | -10.20 h |
See it on a live desk
Predicted ETAs, congestion and sanctions screening across 51 ports — try it in one click, no signup.
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.