Who Tracks Ships in the Strait of Hormuz? AIS, Dark Fleet, and Why Counts Lag 1–4 Days
Explainer · Published 23 Oct 2026 · The public-facing twin of the methodology page
Nobody counts Hormuz traffic by standing on the shore. Every published number is a reconstruction from ship transponders, satellites and radio, and on 23 October 2026 the reconstructions disagree by an order of magnitude: Windward logged 13 transits on 22 October, hormuz.now’s AIS-only line saw 4 crossings in its last 24 hours, and IMF PortWatch reported 4 for 18 October. The ships are real; the difference is who can see a vessel that has switched its transponder off. This page explains who the trackers are, what each one can and cannot see, and why a reliable daily count is always one to four days old.
Status as of 13 Nov 2026: Closed, Day 258. 10 transits on 12 Nov 2026 vs ~88/day pre-crisis. Brent $95.18. Full SITREP →What is AIS, and why is it not enough?
The Automatic Identification System is a VHF transponder that every merchant ship over 300 gross tons must carry under the SOLAS convention. It broadcasts the ship’s identity, position, course and speed every few seconds; terrestrial receivers pick it up near the coast and satellites pick it up at sea. Almost every free “live map” of the Strait of Hormuz, and most of the headline counts in the news, are AIS and nothing else.
AIS has two weaknesses that the 2026 crisis has turned into the whole story. First, the master can switch it off, and a ship that is being shot at has every reason to. Second, the GPS signal that feeds the transponder can be jammed or spoofed: Windward recorded 293 vessels jammed in the Persian Gulf on 21 October alone, and 89,364 unique vessels jammed since 28 February, producing more than three million false “ship-to-ship meetings” in the data (Windward Hormuz Insights). A tracker that only reads AIS therefore undercounts dark ships and can overcount phantom ones in the same hour.
Who are the trackers, and what does each one see?
| Tracker | What it counts | Sees dark ships? | Typical lag | Used on this site |
|---|---|---|---|---|
| Windward | Vessel-level transits, inbound/outbound, AIS plus satellite (SAR and optical) and radio-frequency detection; its MIOC analysts attribute dark transits and keep the threat-zone incident log (106 incidents as of 21 Oct). | Yes, with a lag; some dark attributions arrive days later and revise a day’s count up. | 1–2 days | Headline transit count, inbound/outbound split, incident log, GPS-jamming figures. |
| Kpler | Crossings classified by route (IMO lane, Iranian corridor, dark/unknown) and cargo flows; counted 895 crossings 1 Mar–19 May, 40.6% on dark or unknown routing. | Partly: it classifies dark routing but relies on AIS gaps and cargo data rather than imagery. | 1–3 days | Context only (cited in SITREPs via press). |
| Vortexa | Barrels, not hulls: crude and product outflow by terminal loading and ship-to-ship transfer. | Yes, by reconciling cargo rather than tracks. | 2–3 days | Oil-flow line in every SITREP (via Windward). |
| Lloyd’s List Intelligence | Cargo-carrying vessels over 10,000 dwt, excluding Iranian-linked ships; the attack tally the wires quote (28 attacks in October). | Limited; AIS-based with analyst review. | 1 day for counts, same day for attacks | Attack counts; pre-crisis baseline cross-check. |
| IMF PortWatch | Daily transit calls from satellite AIS for tankers, container, dry bulk, general cargo and ro-ro ships; the free public baseline most fact-checks use. | No; AIS-broadcasting crossings only. | 3–4 days | Baseline cross-check only. |
| MarineTraffic and similar live maps | Terrestrial and satellite AIS positions, shown in near real time. | No. | Minutes | Not used for counts. |
| TankerTrackers | Analyst-run tanker tracking using satellite imagery to spot loadings and dark tankers; strongest on Iranian and shadow-fleet movements. | Yes, by image. | Days, irregular | Not used for counts. |
| Starboard Maritime Intelligence, Marisks, AXSMarine | Analytics firms whose figures appear in Reuters, Bloomberg and Wikipedia; Starboard and Marisks supply traffic counts, AXSMarine dark-transit and trapped-fleet estimates. | Varies by firm. | 1–3 days | Cited where the press cites them. |
| hormuz.now, straits.live | Independent sites; hormuz.now counts AIS-visible commercial ships of 80 m or more crossing its own line, straits.live republishes PortWatch plus prices and holding counts. | No (both say so). | Live to 4 days | straits.live for Brent, war-risk premiums, anchored vessels. |
What is the “dark fleet” in Hormuz?
Three different things get called dark, and they need separating. The first is the pre-war shadow fleet: old, under-insured tankers that moved sanctioned Iranian, Russian and Venezuelan oil with spoofed identities long before February. Kpler found that of 622 unique vessels crossing the strait between March and May, 58.7% were 20 years or older and 55.5% sat outside International Group P&I cover. The second is ordinary commercial ships going dark for safety: Windward estimated in July that around 40% of all Hormuz traffic had switched AIS off, and AXSMarine reported AIS-off transits “nearly doubling” in days when attacks resumed (The Maritime Executive, 10 Jul 2026). The third is the organised night shuttle: tankers moving under US oversight with transponders and phones off, through a protected lane near Oman, to transfer oil ship-to-ship in the Gulf of Oman (AP, 22 Oct 2026). On 22 October Windward’s inbound count was 6 AIS-visible, 1 dark and 2 with no signal at all, which is a normal day’s mix.
Dark does not mean invisible. Synthetic-aperture radar sees steel through cloud and darkness; optical satellites confirm a hull when the sky is clear; radio-frequency sensors locate a ship by the emissions of its own radar and communications; and coastal radar, naval sensors and nearby ships fill in the gaps (The DeepDraft, 31 Aug 2026). But each of those sources sees different ships on different passes, which is why the fused count can only be finished after the day is over, and why it sometimes changes afterwards.
Why does the count lag one to four days?
Four delays stack on top of each other. The UTC day has to close before a transit count for it can exist, so the earliest any provider can publish “yesterday” is the following morning. Satellite passes are scheduled, not continuous, and an image collected overnight is processed the next day; Windward’s own chart notes that a change in its dark-transit series on 28 July reflected “added detection coverage, not a genuine rise in dark activity”. Analyst attribution then takes time: deciding that a radar blob 40 nautical miles off Khasab is the same Panama-flagged product tanker that stopped transmitting off Fujairah is a human judgement, and Windward’s SITREP-day edition carries the last complete day it is confident about, which on 23 October was 22 October. PortWatch, which aggregates satellite AIS globally, publishes about three to four days behind. That is why every SITREP on this site states a data date separately from its publication date, and why the dataset records both.
How does this site choose its number?
The headline is always the Windward vessel-level count for the last complete UTC day, because it is the only widely published series that includes attributed dark transits. The oil-flow line is Vortexa, because barrels and hulls have diverged since the summer: CENTCOM’s statement that “95 percent of vessels are getting through” and Windward’s 13 transits are both reported, and the methodology explains why they measure different things. Iranian, US and company claims about individual ships are reported as claims until a second source confirms them, which is the rule that governs the incidents log.
The site itself is a small open-source intelligence product: it is published by MOCEAN Strategic Cloud, whose cloud and OSINT work is what the daily pipeline runs on, and the pipeline turns each published SITREP into the CSV, JSON and schema outputs described on the data page. No ship-tracking provider funds or reviews the site.
Frequently asked questions
Why do different trackers give different Strait of Hormuz ship counts?
Because they see different ships. AIS-only trackers such as IMF PortWatch, hormuz.now and live maps count only vessels that are broadcasting; Windward and the analytics firms add satellite radar, optical imagery and radio-frequency detection to attribute dark transits. On 27 September 2026, for example, Windward recorded 24 transits where PortWatch recorded 1.
Can a ship really transit the Strait of Hormuz with its AIS off?
Yes, and in 2026 many do: Windward estimated around 40% of Hormuz traffic was dark in July, and Kpler classed 40.6% of crossings between March and May as dark or unknown routing. Switching off AIS breaches SOLAS in normal conditions but is permitted where the master judges that broadcasting endangers the ship.
How old is the transit number in today’s SITREP?
Usually one to two days, occasionally up to four. A count needs a complete UTC day, satellite processing and analyst attribution before it is published. Each SITREP states the data date next to the figure.