Origin & destinationCalled en routeRecorded fixesVessel

Operations over time

Three measures of the same selection: how long the voyage took, how many ships were on the lane, and how much slot capacity they carried. Every chart follows the filters above. Turn a legend on to split that chart by operator.

Period

Voyage time

Legend off: each column is the distribution — the bright rule is the median, the block the middle half, the spine the fast and slow tenth. Legend on: one line per operator through its median. The axis starts near the data, not at zero; it is a distribution comparison, not a magnitude one.

Split by operator

Vessels deployed

Distinct ships that sailed the lane in each period. A carrier adding tonnage shows up here before it shows up in transit time.

Split by operator

Capacity deployed

Slot capacity carried — the TEU of every sailing in the period, summed. Not cargo moved: it is the space offered.

Split by operator

Services on this lane

Each operator’s busiest strings, most voyages first. Open a service to see the ports its ships actually called, how long it takes and which vessels run it — all measured over the current selection. Service names are labels the carriers publish, so treat them as a handle on the rotation, not an identifier.

Shipping lines on this lane

One card per operator, biggest share of voyages first — the order re-ranks as you filter or scrub the timelapse. Every figure is measured over the current selection only. Vessel particulars are the largest and typical ships each line actually used here.

Voyages

One row per paired voyage on this lane: the ship left Shanghai, arrived Savannah, and the record says when, with which terminals at each end. The filter bar above scopes this table like every other pane; the filters in the header row narrow it further.

Container records

Carrier container data. Direct means the box was never transshipped — it stayed on the vessel it first loaded onto. Via a port means it was transshipped there and carried on aboard a different vessel. This tier is here to establish one fact: that containers really do move Shanghai → Savannah. Picking operators above scopes the table where the carrier appears in both feeds.

What this page measures

Shanghai → Savannah is measured twice here, from two independent records. Carrier container tracking records 6 routings on this lane (60 carrier/service combinations, 13 operators); sailing direct, a container typically takes 36.58 days gate to gate over 491 recorded voyages — the same figures the route finder shows for this pair. AIS ship movements give 1,524 port-to-port voyages by 421 vessels, median 38.7 days (mean 41.2 d, middle half 34.8–44.7 d).

The two medians are not comparable directly: 2 carriers record containers on this lane against 17 operators seen by AIS, so the two sets cover different ships. Where they can be compared — one carrier at a time — the difference is port time, measured at 3.53 d at Shanghai (12,464 records) and 3.45 d at Savannah (7,091 records).

Of the 381 journeys whose routing is recorded, 358 moved without transshipment and 23 were transshipped — Busan (20, median 43.4 d); Shanghai (3, median 31.8 d) .

The ships do not sail this lane in a straight line. The most called intermediate port is Busan, on 76% of voyages (1,153 of 1,524), and the median voyage makes 3 intermediate calls (the most recorded is 7). A container the carrier records as direct still rides a ship that stops along the way — direct means no transshipment, not nonstop.

Voyage time is not stable. By year the median ran from 36.6 d in 2026 to 45.9 d in 2022, a spread of 9.3 d across 5 years.

It is not uniform across carriers either. Over the whole period the quickest well-evidenced service is TP11 at 31.8 d (Maersk, n=55) and the slowest of the ranked set is AWE4 at 39.5 d (Cosco Shipping Lines, n=73) — a difference of 7.7 d on the same lane.

Nothing here is scheduled, modelled or estimated. Every figure is a count of observed movements, and every figure carries that count.

Where these numbers come from

Two record sets, kept separate: AIS ship movements and carrier container records. Every figure on this page is measured from the same files you can export below, and each one is listed here with the expression that regenerates it.

1 · Reproduce every figure

Each row names a figure on this page, what it means, and the exact expression that regenerates it from the exported file. Column names are the CSV’s own. vessel-calls.csv is one row per paired voyage; container-records.csv is one row per container journey. Both export the whole lane, unfiltered — apply the same date or operator filter you have on screen to reproduce a filtered figure.

Figure	Value	What it means	Reproduce it
Paired voyages	1,524	One ship leaving Shanghai and reaching Rotterdam. The unit behind every AIS figure on the page.	vessel-calls.csv · `COUNT(*)`
Distinct vessels on this lane	421	Individual hulls, however many times each sailed.	vessel-calls.csv · `COUNT(DISTINCT imo)`
Ship median transit	38.7 d	Middle voyage, port to port. Hull in water — it excludes nothing and includes any intermediate calls.	vessel-calls.csv · `MEDIAN(transit_days)`
A shipping line’s sailings	152 for Maersk AE5	Voyages that carrier ran on that one service string.	vessel-calls.csv · `COUNT(*) WHERE operator=… AND service=…`
A shipping line’s median	31.9 d for Maersk AE5	Middle voyage of that operator + service pair.	vessel-calls.csv · `MEDIAN(transit_days) WHERE operator=… AND service=…`
Vessels deployed, per period	Operations pane	Distinct hulls that sailed in the period.	vessel-calls.csv · `COUNT(DISTINCT imo) GROUP BY year(departure_utc)`
Capacity deployed, per period	Operations pane	Slot capacity offered: the TEU of every sailing, summed. Not cargo moved.	vessel-calls.csv · `SUM(teu) GROUP BY year(departure_utc)`
Ports called en route	map dots	Share of voyages that called there. A voyage counts once per port however often it called.	vessel-calls.csv · `COUNT() WHERE ports_en_route LIKE '%port name%' / COUNT()`
Tracked journeys on this lane	397	Containers whose whole door-to-door journey the carrier published.	container-records.csv · `COUNT(*)`
Median door to door	33.0 d	Middle container journey, including port time at both ends and at any hub. Not comparable with the ship median above.	container-records.csv · `MEDIAN(transit_days)`
Middle half	30.9–34.7 d	The 25th to 75th percentile of those journeys.	container-records.csv · `QUANTILE(transit_days, .25) … (.75)`
Carriers recording this lane	2	Carriers publishing container events here — fewer than the operators AIS sees.	container-records.csv · `COUNT(DISTINCT carrier)`
Direct vs via	Containers pane	Direct means never transshipped. Via names the hub where the box changed vessel.	container-records.csv · `GROUP BY transship_ports` (empty = direct)

Two grains, stated plainly. The routing cards in the Containers pane follow the route finder: every figure there is a median over the voyage records that route-report (the Export data workbook) rebuilds row by row. container-records.csv is a different cut — the all-time booking log at one row per door-to-door journey (397 rows here), with no recency window. Use the workbook to reproduce the cards; use the CSV for the raw journey log.

2 · Export this data

Both record sets download as CSV at their own grain. Container records are the routing/carrier rows behind the Containers pane; vessel calls are one row per paired voyage with its particulars and the ports it called. Row allowances are per account and per calendar month.

Checking your allowance…

Container records

—

one row per container journey on this route: carrier, container number, transshipment ports, departure and arrival, legs and days

Vessel calls

—

one row per paired voyage: vessel, IMO, TEU, LOA, draught, both terminals, hours alongside, ports en route