Calculating ETA to the Boarding Point — and When to Launch the Pilot Boat

Every pilot boat run that leaves the dock too early burns fuel and crew time idling at the boarding ground. Every run that leaves too late puts a deep-draft ship in a holding pattern off the sea buoy, sometimes in a current and a tide window that won't wait. The whole problem sits in one number: when, exactly, does the inbound vessel reach the boarding point — and therefore, when does the boat need to be in the water to meet it?

That number looks simple. It is distance over speed. But the speed lies, the distance has to be measured along the right line, and the answer is only useful if it triggers an action at the right minute. This is the mechanics of the launch decision — the part of dispatch that a board built for pilots, like Binnacle Passage, is supposed to get right so a human doesn't have to do it in their head at 0300.

Step one: the distance, measured honestly

The inbound vessel is a position broadcasting over AIS — a latitude and longitude updating every few seconds. The boarding point is a fixed coordinate: the pilot boarding area, often at or near the sea buoy or a charted boarding ground. The first quantity you need is the distance between them.

For the leg of open water between an offshore vessel and a boarding ground, the right tool is the great-circle distance — the haversine of the two coordinates on a sphere. Over the tens of miles that matter here, the curvature correction is small but real, and using a flat-earth approximation introduces error you don't need to carry. The board computes great-circle range continuously as the vessel's position updates.

A caveat worth stating plainly: great-circle range is a straight line over the ground. If the approach bends around a shoal, a traffic separation scheme, or a dogleg in the fairway, the vessel's actual track is longer than the rhumb between the two points. A serious board lets you measure along the approach route, not just point-to-point, so the distance reflects the water the ship will actually transit. Point-to-point is a fine first cut offshore; it gets worse the closer the vessel is to a confined approach.

Step two: the speed that lies

Divide distance by speed and you have an ETA. The speed you get from AIS is speed over ground (SOG) — the vessel's velocity relative to the seabed, derived from GPS. That is exactly the speed you want for an ETA, because the ground is what the distance is measured over.

Here is where the trap is. SOG is a snapshot. The current at the vessel's present position is not the current it will see for the rest of the run. A ship making 12 knots SOG into a 2-knot ebb is doing 14 through the water; when that ebb slackens and turns to flood over the next ninety minutes, the same 14 knots through the water becomes 16 SOG. Your ETA, computed off the instantaneous 12, is now wrong — and wrong in the direction that matters, because the ship arrives early and the boat isn't there.

This is not a rounding error. In a district with a strong reversing tidal stream, the along-track current component can swing several knots across a single approach. A speed-over-ground ETA that treats the current as frozen can be off by a meaningful chunk of an hour over a long run. We've written about exactly this dynamic in the hardest version of the problem — see Cook Inlet pilotage dispatch, where the tidal range is among the largest in North America and a tide-blind ETA is nearly useless.

There are two honest ways to handle it:

• Use a smoothed SOG rather than the raw instantaneous value, so a momentary yaw or a single noisy GPS fix doesn't whip the ETA around. A short rolling average over the last several position reports is enough to settle it.
• Fold predicted current into the run. NOAA's CO-OPS predictions give the current at stations along the approach; projecting the along-track component onto the remaining distance produces an ETA that anticipates the turn of the tide instead of being surprised by it.

The first is table stakes. The second is what separates a board built for pilotage from a generic AIS plot that just draws a dotted line ahead of the icon.

Step three: the transit time you're racing against

The ETA tells you when the ship arrives. It says nothing about when the boat has to leave. That's a separate, district-specific number: the station-to-boarding transit time — how long it takes the pilot boat to run from its berth or station to the boarding ground.

This is a property of your operation, not the ship's. It depends on the boat, the distance from the pilot station to the boarding area, the sea state, and any speed restrictions in the harbor. Some districts measure it in single-digit minutes; a station far up a river or across a bar can be 30, 40, or more. On the Columbia River Bar, the boat run and the bar conditions are a first-order part of the timing, not an afterthought. The point is that the transit time is a known quantity for your station — and once it's known, the launch math is trivial.

Step four: the amber launch alert

Put the two numbers together and the launch decision is arithmetic:

Launch when ETA-to-boarding minus station-to-boarding transit time drops to or below your lead buffer.

In practice the board carries a small lead buffer on top of raw transit time — a cushion of a few minutes so the boat is waiting for the ship, not the other way around. A common shape: when ETA − transit ≤ 15 min, the vessel's row goes amber and the dispatcher gets a launch alert. That 15-minute threshold is configurable per district, because a 6-minute boat run and a 35-minute boat run want very different windows.

The alert is the whole point. The dispatcher isn't recomputing haversine in their head; they're watching a board that turns amber at the right moment and tells them: this is the one, send the boat now. The ETA can be live and current-aware underneath, but the output the human needs is a single, timely, unambiguous cue.

What that cue saves, concretely:

• A wasted run. Launch on a tide-blind ETA that has the ship arriving 25 minutes early, and the boat sits at the boarding ground burning fuel and crew hours while the ship is still inbound against a current that's about to slacken. Get the timing right and the boat leaves once.
• A ship waiting. Launch late and a laden ship slows or circles off the entrance — costly for the vessel, and worse if the boarding window is tied to a tide or a daylight constraint that's closing.
• A blown tide window. For deep-draft transits, the boarding has to line up with the under-keel-clearance window, not just the boat's availability. The launch alert has to respect the tide, which is exactly why the ETA and the tide window belong on the same screen.

Why the tide window sits next to the ETA

The boarding time is rarely free to be whatever the math returns. For a deep-draft inbound, the transit through the channel has to fall inside the window where there's enough water under the keel — and that window is set by the tide, not the boat schedule. If the launch alert fires for a 0410 boarding but the UKC window doesn't open until 0450, the right answer isn't to launch; it's to hold and re-time.

So the ETA logic and the tide logic are two halves of the same decision. Our free tide-window calculator does the standalone version of this math — given draft, channel depth, and the tide curve, when is the passage actually clear? On a live board, that window is overlaid against the inbound ETAs so the launch alert only fires when both conditions agree. (For the squat side of the UKC question, the squat calculator covers how speed eats into your clearance.) And once the transit is complete, the same record that timed the boat feeds the pilotage tariff calculator and the invoice — the ETA, the boarding, the standby clock if the ship was late, all tied to one transit.

The judgment the board protects

None of this removes the dispatcher or the boat operator from the loop. A 2-knot current prediction is a prediction; fog, traffic, a balky main engine, or a master who slows for his own reasons can all move the real arrival. The boat crew reads the sea state the model can't feel. What the board does is remove the arithmetic burden — the haversine, the current projection, the transit-time subtraction, the watching-the-clock — so the human attention goes to the judgment calls that actually need a human.

That's the difference between a vessel-tracking tool that happens to show ships and a dispatch board built for the launch decision. One draws an icon and a dotted line. The other measures the distance along the real approach, fronts a current-aware ETA, knows your station's transit time, and turns the row amber at the minute the boat should leave — defensibly, every time, on the night watch as reliably as the day.

See it on a live board

The launch alert, the current-aware ETA, and the tide-window overlay are the core of what Binnacle Passage does for a US pilots association. The fastest way to understand it is to watch a board run — see a live pilot board and walk an inbound vessel from offshore, through the amber launch alert, to a timed boarding.

This article is general information and not operational guidance. ETAs and launch timing are the responsibility of the dispatcher, the pilot, and the boat operator, governed by the licensed pilots association serving the district.