Pilot Fatigue and Rest-Hours: Keeping an Unfit Pilot Out of the Rotation

A pilot association does not have a fleet to manage. It has a body of water, a duty roster, and a rotation — and the single most consequential decision the dispatch desk makes all night is which name is next. Most of the time that decision is about equity: whose turn is it, who is owed a job, who has been short. But there is a second question riding underneath every assignment, and it is the one the National Transportation Safety Board keeps writing about in its incident reports.

Is the next pilot actually fit to take the ship?

Fatigue is repeatedly cited as a contributing or causal factor in groundings, allisions, and contact casualties involving compulsory pilotage. A pilot who has been awake for twenty hours, who came off a six-hour upbound transit four hours ago and is now being woken to take a deep-draft tanker out on a falling tide, is not the same pilot who reported for duty fresh. The board knows this. The pilots know it. The hard part has always been operationalizing it — turning "don't send a tired pilot" into a rule the dispatch desk can apply at 0300 without an argument.

Why fatigue is the dispatch desk's problem, not just the pilot's

There is a comfortable assumption that fitness for duty is the individual pilot's call. Professionally, it is — a licensed pilot can and should decline a job. But operationally, the desk owns the consequences. The dispatcher assigns the work, the association carries the liability, and "the pilot should have said no" is a thin defense when the transit data shows the desk handed a job to someone who had been on the water most of the night.

Pilots board at all hours. Traffic does not respect a day shift. A busy district will run inbound and outbound moves across the full 24, and the boarding times are dictated by tide, by berth availability, by the ship's own schedule — none of which care whether the pilot next in rotation slept. That is exactly the environment in which fatigue accumulates invisibly: no one is tracking cumulative hours-on-task because the rotation is tracking turns, not hours.

The rest-hour frameworks the rest of the maritime world runs on — STCW and the Maritime Labour Convention's minimum-rest rules (10 hours in any 24, 77 in any 7-day window, rest split into no more than two periods) — were written for crew on articles, not for pilots who board, transit, and disembark. Pilots sit in an awkward gap. Compulsory state pilotage under 46 USC 8501 is governed state by state, and many state commissions impose their own rest requirements between assignments. But a paper board and a VHF radio give the dispatcher no running tally of who is approaching a limit. The rule exists; the enforcement mechanism does not.

Deriving work hours from the transit, not from a sign-in sheet

The first thing a modern dispatch board should do is stop trusting self-reported hours and start deriving them from what actually happened on the water.

When the association runs an AIS-backed board, every assignment already produces a transit record: the pilot boarded vessel X at the boarding ground at time T1, the vessel transited the district, and the pilot disembarked at T2. That is a measured work period, not an estimate. The same position data that feeds the pilotage tariff calculator and the per-GT billing line is the cleanest possible source of truth for hours-on-task — because it is the ship's own broadcast, time-stamped, not a number someone scribbled on a clipboard after the fact.

From a stream of those records, a board can compute the three numbers that actually drive a fatigue rule:

• Time since last disembark — the continuous rest the pilot has had since the last job ended. This is the gate for the next assignment.
• Cumulative work in the trailing 24 hours — the sum of transit durations (plus a configurable boarding/transit-out overhead) across the rolling day.
• Consecutive jobs without a qualifying rest period — the count that catches a pilot getting nibbled to death by short moves that each look harmless in isolation.

None of those require the pilot to log anything. They fall out of the transit data the board is already keeping for billing and for the audit trail. That is the quiet power of deriving hours from transits: the safety number and the revenue number come from the same measured event, so neither can drift from reality.

The two rules worth enforcing

You do not need a complicated fatigue model to make a real safety difference. Two rules, configurable to match your state commission's requirements, cover most of the risk:

1. Minimum continuous rest before the next job. A pilot must have had at least N hours off the water — measured from last disembark to proposed boarding — before being eligible for the next assignment. Set N to whatever your commission mandates, or to the association's own standard if it is stricter.
2. Maximum work in a rolling 24 hours. Once a pilot's summed transit time in the trailing day crosses the ceiling, they are ineligible until enough of that window rolls off.

These are deliberately simple. The point is not to model sleep science; it is to put a hard floor under the worst decisions — the ones that look fine because the rotation says it is someone's turn, and no one is watching the clock that matters.

The rotation that skips an unfit pilot automatically

Here is where it comes together. A rotation board that knows each pilot's derived hours can do something a paper board never could: it can take the next-up pilot out of the running automatically when they fail a rest gate, and present the next eligible name — without breaking rotation equity.

That last clause is the whole game. If skipping a tired pilot meant they lost their turn, the system would quietly punish people for being assigned hard nights, and the pilots would learn to defeat it. A board built for pilots holds the skipped pilot's position. They are passed over for this job because they are inside a rest window, but they remain owed the turn — the equity ledger that governs Cook Inlet's brutal all-hours rotation and every other district is preserved. The fatigue gate changes who takes this ship; it does not change who is owed work over the month.

Mechanically, when the dispatcher goes to assign the next inbound, the board:

• Looks at the rotation order.
• Checks each candidate against the rest rules using their derived hours.
• Flags the next-up pilot as resting with the specific reason ("4.2 h since disembark, needs 8") and the time they become eligible.
• Offers the next eligible pilot, while holding the resting pilot's place in line.

The dispatcher still has authority — a documented override exists for genuine operational necessity, because real life happens. But the default is the safe choice, the override is logged, and the reason is captured. That inversion — safe by default, exceptions on the record — is exactly what an after-incident review wants to see.

The liability case, stated plainly

Consider the two versions of the same night in front of an investigator or a court.

Version one: the desk ran a paper board. There is no record of the assigned pilot's hours-on-task. The transit times exist only in scattered logs. When asked "did you know this pilot had worked eighteen of the last twenty-four hours when you assigned the outbound," the honest answer is "we didn't track that."

Version two: the board derived the pilot's hours from AIS transit records, applied the rest rule, flagged the pilot as inside a rest window, and the dispatcher made a documented override for a specific reason. Every step is time-stamped and attributable.

Version two is not a guarantee against an incident — nothing is. But it transforms the association's posture from we had no system to we had a system, it worked, and a deliberate decision was made and recorded. That is the difference between negligence and a defensible operating procedure. The same audit trail that protects the per-GT tariff line and answers a billing dispute is the trail that demonstrates the association took fitness-for-duty seriously.

And the safety upside is the real point. The grounding that does not happen, the allision that does not bend a dock, the contact casualty avoided because a fatigued pilot was held for four more hours of rest — those are the wins that never make a headline. A fatigue gate is cheap insurance against the kind of casualty that ends careers and, sometimes, lives.

Where this sits in the rest of the board

Fatigue management is not a bolt-on; it is one facet of a board that already knows the district's traffic, the tide, and the under-keel margins. The same system that overlays NOAA current onto the ETA and computes squat for a deep-draft transit — the math behind the tide-window calculator and the squat calculator — is the natural home for the rest ledger, because it is already holding the timeline of every transit. Binnacle Passage treats the rotation, the transit record, and the fatigue gate as one connected object rather than three disconnected tools, which is the only way the safety number and the equity ledger stay honest with each other.

The pilots already do the hard part on the water, in the dark, in weather, at all hours. The desk's job is to make sure the person climbing the ladder is fit to be there — and to have the record to prove it.

If your association is still tracking the rotation on a whiteboard and trusting memory for hours-on-task, you are carrying a risk you do not have to. See how a derived-hours rotation with rest gates works in Binnacle Passage, or see a live pilot board and watch a fatigued pilot get skipped — and held — in real time.

This article is general information and not legal advice. Rest-hour requirements for compulsory pilotage are set by individual state pilot commissions under 46 USC 8501; confirm your district's rules with your commission.