Marine Autopilot Guide

On a shorthanded offshore yacht the autopilot can be the hardest-working system aboard. It steers for days, in conditions that would exhaust a human helm, and its failure converts a comfortable passage into an ordeal.

That makes it something quite different from the convenience feature it is on a day boat, and it explains why the most important part of the decision is not the part with the buttons on it.

#The control loop

Desired heading
      |
      v
Autopilot computer <---- heading / motion sensor
      |
      v
Drive
      |
      v
Steering
      |
      v
Boat turns
      |
      +---- feedback

Four things have to be right: the sensor must know where the boat is pointing, the computer must decide sensibly, the drive must be able to move the rudder, and the feedback must close the loop. A weakness anywhere degrades the whole thing, and the weakness is rarely in the computer.

#Main components

• heading and motion sensor — an AHRS or GNSS compass
• pilot computer — the control logic
• drive — the muscle that moves the rudder
• control head — the interface, and the least important part
• rudder reference where fitted, reporting actual rudder angle
• NMEA network, carrying it all
• wind and navigation inputs for sailing and route modes

#Drive sizing: the decision

Never size an autopilot only by boat length. This is the error that produces most disappointing installations, and it persists because length is the number in every brochure.

What actually determines the load:

• displacement, and how the boat is loaded for cruising
• rudder area
• rudder balance — how much of the area sits forward of the stock
• steering geometry, and the mechanical advantage it provides
• hydraulic cylinder volume where applicable, which sets how much fluid the pump must move
• maximum steering load in the worst realistic conditions you will meet

A heavy 40 ft cruiser can require far more drive force than a light 40 ft racer. Same length, different machine.

And read a manufacturer's published displacement limits as a specification, not a guideline. A drive at the edge of its envelope in a quartering sea is a drive that will be replaced, probably at an inconvenient time.

#Drive types

The drive is the muscle. An excellent pilot computer cannot compensate for an undersized drive, a badly mounted one, or a mounting bracket that flexes under load. Inspect the mechanical installation as carefully as you choose the electronics — see marine electronics refit.

#Heading quality

Autopilot performance depends on accurate motion sensing, and poor sensor placement degrades steering in ways that look like a control problem.

Keep magnetic sensors away from loudspeakers, high-current cables, motors, alternators, batteries, steel objects and magnets. Intermittent fields — a windlass feed, a thruster — are worse than constant ones, because calibration cannot model something that was absent when you calibrated.

A pilot fed by a well-sited sensor with rate-of-turn output can anticipate rather than merely react. One fed by a sensor beside a speaker will hunt, and no amount of gain tuning fixes it. The full argument is in heading sensors.

#Wind steering

A sailing pilot can steer to apparent or true wind angle, which offshore is often the right mode: the boat follows the wind's shifts rather than fighting them onto a fixed compass course.

It works brilliantly, and it requires accurate inputs:

• wind, from a correctly aligned masthead unit
• heading, to derive true wind at all
• boat speed through water, where the calculation uses it

Bad input produces bad steering, and the failure is often misread as a pilot fault. A wind sensor with an uncalibrated alignment offset gives a constant apparent-wind-angle error, which the pilot faithfully steers to.

#Sail balance is an electrical question

An overpowered yacht with heavy weather helm forces the pilot to work harder against a rudder that is fighting it.

The consequences are cumulative:

• higher electrical consumption, sometimes dramatically
• faster mechanical wear on the drive and linkage
• poorer steering, because the pilot is at the edge of its authority

Reefing earlier is an autopilot efficiency measure as much as a seamanship one. On a boat where the pilot dominates the energy budget, it is also a power-management measure.

#Power consumption

The pilot load varies enormously with conditions, and the variation is the point.

A calm passage may need only small, occasional corrections. A quartering sea may demand near-continuous drive operation, and on many cruising yachts that is the largest continuous consumer aboard — capable of drawing ten times its calm-water figure.

So measure real current use on your own boat, in realistic conditions, and put that number in the 24-hour energy budget. Manufacturer figures describe the computer, not the drive working hard in a seaway. This is also why satellite broadband and the autopilot are competing for the same amp-hours — see marine DC power.

#The ecosystems

#B&G

Particularly strong in sailing autopilots. H5000 provides advanced sailing-oriented control behaviour for performance and bluewater boats; NAC products suit mainstream cruising installations. See B&G marine electronics.

#Raymarine

Evolution systems are widely used on cruising yachts and integrate tightly with Axiom displays and Raymarine instruments. Well proven in exactly the shorthanded offshore role that matters most. See Raymarine marine electronics.

#Garmin

Reactor systems cover power and sailing applications. In September 2026 Garmin announced SmartDrive for compatible sailboats and catamarans — a brushless linear actuator with integrated ECU and rudder feedback, with stated compatibility for boats up to 45 ft within published displacement limits.

That makes Garmin a materially more serious sailing-autopilot competitor than it was, and it should now be compared directly against B&G and Raymarine rather than dismissed. See Garmin marine electronics.

#Simrad and Furuno

Both cover powerboat and commercial-derived applications, with Simrad strong on hydraulic installations for motor yachts and sportfishers.

#Offshore redundancy

Hand-steering for days is a safety problem, not an inconvenience. Possible backup approaches:

• a windvane — the strongest option for a sailing yacht, because it shares no failure domain with the electrical system
• a second drive, where the linkage and space allow
• a spare actuator or motor, carried and fitted once in practice so you know it fits
• a tillerpilot on suitable boats
• the emergency tiller — located, tested, and reachable with the cockpit full of gear

The correct strategy depends on the actual steering system, and the useful question is which single failure leaves you steering by hand for the rest of the passage. See navigation redundancy.

#Commissioning

A pilot is not commissioned when it holds a course in flat water. Test:

• drive direction — a reversed drive is dangerous and easy to create
• rudder limits, with mechanical clearance confirmed
• heading calibration, and which source the network selected
• sea-state response, including rudder gain and counter-rudder
• behaviour in a following and quartering sea, which is where an undersized or badly tuned pilot reveals itself
• tack and gybe functions, on both tacks
• route following, and that it prompts rather than turning autonomously at waypoints
• instant disengagement under load — and make sure the crew can do it
• actual current draw, in calm and in a seaway

Do not undertake a major passage immediately after installation without a sea trial. Autopilot behaviour in real conditions cannot be verified at the dock, and the quartering-sea test is the one that finds the problems. See commissioning and sea trials.

#Bottom line

The right autopilot is the system that can physically control the rudder in the worst realistic operating conditions, with enough margin and enough redundancy for the voyage. Size the drive from the steering system, site the heading sensor for physics, measure the real current draw, and sea-trial it in a following sea before you rely on it.