ETA accuracy benchmark · 🇯🇵 Yokohama / Tokyo Bay
How Port Flow's predicted ETA compares to each vessel's own broadcast ETA at Yokohama / Tokyo Bay, 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)
11.87 h
MAE 17.22 h · 330 voyages
Broadcast ETA (RMSE)
53.70 h
crew-declared, 125 voyages (same set)
Our advantage
77.9 %
lower error than the broadcast ETA
Broadcast ETAs more than 7 days from the actual arrival are excluded as AIS sentinel / placeholder values (67 excluded). Both figures are computed on the same remaining voyage set — apples to apples.
Closed voyages
error = predicted/broadcast minus actual arrival (hours)| MMSI | Cargo | Arrival | Our error | Broadcast error |
|---|---|---|---|---|
| 431010468 | container | 07/12, 11:25 AM UTC | 0.84 h | -915.42 h |
| 265491000 | container | 07/12, 09:31 AM UTC | 1.40 h | -2.37 h |
| 563316400 | container | 07/12, 08:55 AM UTC | 3.12 h | -432.93 h |
| 431005246 | container | 07/12, 08:09 AM UTC | -46.09 h | -372.15 h |
| 440774000 | product | 07/12, 07:39 AM UTC | 0.14 h | -10.16 h |
| 431501701 | product | 07/12, 06:55 AM UTC | -7.89 h | -73.93 h |
| 215197000 | container | 07/12, 06:40 AM UTC | 0.31 h | -2.68 h |
| 431028356 | container | 07/12, 06:39 AM UTC | 0.05 h | -58.15 h |
| 431100969 | product | 07/12, 05:42 AM UTC | 0.80 h | -6.55 h |
| 431019444 | product | 07/12, 05:33 AM UTC | -22.86 h | — |
| 431004235 | product | 07/12, 05:27 AM UTC | 5.88 h | 9.05 h |
| 431600647 | container | 07/12, 04:30 AM UTC | 2.27 h | 9.49 h |
| 431030514 | container | 07/12, 04:07 AM UTC | -3.42 h | — |
| 431018712 | product | 07/12, 03:20 AM UTC | 4.20 h | 5.33 h |
| 563116500 | general-cargo | 07/12, 02:51 AM UTC | -1.52 h | -34.61 h |
| 431018825 | product | 07/12, 02:27 AM UTC | -0.38 h | — |
| 354850000 | product | 07/12, 01:39 AM UTC | 0.30 h | -48.17 h |
| 431027872 | container | 07/12, 01:31 AM UTC | 0.14 h | -710.52 h |
| 563319700 | container | 07/12, 01:23 AM UTC | -25.66 h | -751.39 h |
| 431009748 | container | 07/12, 12:00 AM UTC | -0.89 h | -104.02 h |
| 431005995 | container | 07/12, 12:00 AM UTC | -0.38 h | -641.00 h |
| 431800699 | product | 07/11, 11:47 PM UTC | -1.30 h | -40.79 h |
| 431001071 | product | 07/11, 11:42 PM UTC | 2.29 h | — |
| 311000102 | container | 07/11, 11:38 PM UTC | 1.75 h | -2.06 h |
| 440774000 | product | 07/11, 11:27 PM UTC | 3.00 h | -1.96 h |
| 356688000 | container | 07/11, 11:21 PM UTC | -0.04 h | -2.85 h |
| 431004154 | container | 07/11, 10:58 PM UTC | 0.20 h | 0.03 h |
| 431009162 | product | 07/11, 10:56 PM UTC | 2.86 h | — |
| 431018129 | container | 07/11, 10:51 PM UTC | 2.08 h | -4.86 h |
| 538009273 | container | 07/11, 10:38 PM UTC | 3.43 h | 8.36 h |
| 563487000 | container | 07/11, 10:36 PM UTC | 7.69 h | -410.61 h |
| 636022116 | container | 07/11, 10:35 PM UTC | 2.78 h | 2853.32 h |
| 431011753 | product | 07/11, 10:24 PM UTC | -24.28 h | — |
| 431014183 | container | 07/11, 10:23 PM UTC | 2.08 h | — |
| 431000224 | product | 07/11, 10:23 PM UTC | -0.52 h | 1.61 h |
| 431100938 | product | 07/11, 10:17 PM UTC | -112.03 h | -85.55 h |
| 431101028 | product | 07/11, 10:13 PM UTC | 0.77 h | — |
| 431005016 | product | 07/11, 05:56 PM UTC | -18.65 h | — |
| 548706000 | container | 07/11, 04:50 PM UTC | -50.23 h | -331.17 h |
| 477238300 | container | 07/11, 04:23 PM UTC | 0.25 h | -0.89 h |
| 212334000 | general-cargo | 07/11, 02:26 PM UTC | 4.94 h | -3.20 h |
| 440774000 | product | 07/11, 12:44 PM UTC | -0.23 h | 8.76 h |
| 574012854 | container | 07/11, 12:30 PM UTC | -11.27 h | — |
| 431007373 | container | 07/11, 12:04 PM UTC | 1.64 h | 0.10 h |
| 477238300 | container | 07/11, 11:27 AM UTC | -12.25 h | -324.46 h |
| 431007373 | container | 07/11, 11:15 AM UTC | -14.88 h | -68.93 h |
| 370616000 | container | 07/11, 11:14 AM UTC | 1.27 h | -0.24 h |
| 431004154 | container | 07/11, 11:12 AM UTC | -5.63 h | — |
| 249324000 | container | 07/11, 11:08 AM UTC | 1.22 h | -2.15 h |
| 431004611 | product | 07/11, 10:56 AM UTC | 3.46 h | -6.44 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.