Why Your Chartplotter Throws Away Your Engine Data (And What to Do About It)

Your engine broadcasts roughly fifty telemetry messages per second onto the NMEA 2000 backbone. Your chartplotter renders the latest one as a number on a gauge — and then deletes it. The history that would actually tell you how your engine is aging? It never gets written down.

Fifty messages per second, and your MFD keeps almost none of it

A modern outboard or sterndrive — a Yamaha F300, a Mercury Verado, a Volvo Penta D4 — pushes a steady firehose of NMEA 2000 traffic onto the bus. PGN 127488 (Engine Parameters Rapid Update) carries RPM at roughly 10 Hz. PGN 127489 (Engine Parameters Dynamic) adds coolant temperature, oil pressure, fuel rate, alternator voltage, total engine hours, and a bitfield of warning states once per second. PGN 129026 fires SOG and COG at 10 Hz. Add GNSS position, depth, water temp, and rudder angle and you're well past fifty discrete messages per second.

Your Garmin GPSMAP 8612, your Raymarine Axiom Pro, your Simrad NSS Evo3 — they all see every one of those messages. They have to, to display them. But they're display devices. The number on the screen is the most recent reading; the one before it is gone. The chartplotter is a real-time renderer wearing the costume of a recorder.

What an MFD actually saves to disk

Multifunction displays do persist some data. It's just not the data you'd want for diagnosing your engine. Across the major brands, the storage profile looks roughly like this:

• Waypoints, routes, and tracks. GPS breadcrumbs are saved by default. Garmin GPSMAP units store tens of thousands of trackpoints; Raymarine Axiom and Simrad NSS do similar. This is the original chartplotter use case.
• An alert and event log. When the engine triggers a warning, most MFDs write it to a small alert log — often the last few hundred entries. Useful for "what alarm went off last Saturday," useless for trend analysis.
• Sonar history. Down/side-scan units keep a scrolling sonar log, sometimes recordable to SD card.
• Fuel manager totals. Most MFDs let you enter a tank size and integrate fuel rate over time. The total survives reboots; the second-by-second stream that produced it does not.

What they don't save: the time series. Not RPM-by-second, not fuel rate at 4500 RPM across your last twenty trips, not coolant trend over the season, not engine hours per RPM band, not oil pressure at idle versus WOT. None of it.

Why the MFD vendors made this choice

This isn't laziness or bad faith. Garmin, Raymarine, Simrad, B&G, and Furuno built chartplotters to do navigation. The product brief — going back to the original handheld GPS units — was "show me where I am, where I'm going, and what's under the boat." Engine data showed up later as a courtesy display, not the core job.

Persistent time-series storage is also genuinely hard: you need a database, schema migrations across firmware updates, an export path, and a UI that turns thousands of points into something glanceable while docking. None of that fits the chartplotter form factor, and a chunk of their customers would never use it. So MFDs do what makes sense for a navigation device: render the live value, fire an alert on threshold, move on.

What you lose when nothing keeps the trend

Engines mostly don't fail with a bang. They fail with a slope. The diagnostic signal is almost always in the trend, not the snapshot:

• Fuel rate at a given RPM, drifting up. A 4–8% increase in fuel burn at cruise over a season usually means a fouling prop or hull, a worn impeller, or a dirty injector. Each one costs you money before it costs you a tow.
• RPM-band hours. "Engine hours" on the dash is one number. What matters for service intervals and resale is hours at idle vs. cruise vs. WOT. A 1,000-hour engine that lived at 3500 RPM is a different machine than one that spent half its life at 5800.
• Coolant temperature creep. A coolant temp that used to settle at 165°F at cruise and now settles at 178°F is the early warning of a fouling heat exchanger, weak water pump, or partially blocked raw-water intake.
• Battery resting voltage at engine-off. A bank that read 12.7V at rest in May and reads 12.4V in August is dying — and a chartplotter has no way to know what "at rest" even means without storing context.
• Oil pressure at hot idle. A slow downward drift is bearing wear, often years before the engine knocks.

A chartplotter shows you NOW. None of these are visible in NOW — they're visible across weeks and seasons of NOW.

A worked example: catching an impeller before it cooks the engine

A captain runs the same 30-mile loop most weekends — Yamaha F300, clean prop, hull waxed last month. Eight trips in, the chartplotter looks normal: 4500 RPM at cruise, coolant 162°F, oil 58 PSI. Nothing alarming, nothing logged. But the time-series, if anyone were keeping it, would show this:

That's a 6.1% fuel-rate rise and a 10°F coolant creep across eight trips. None of those numbers crossed an alarm threshold — the Yamaha won't fire overheat until ~205°F. The MFD only registers the failure after the fact: a single alert-log line that says "engine overheat — Sunday afternoon."

The pattern is almost unambiguous: the raw-water impeller is worn or the heat exchanger is fouling. A captain who saw this trend would change the impeller on a Saturday for $30. A captain who only had the chartplotter limps home with a cooked powerhead and a $6,000 invoice.

The serious-boat-owner alternative

The fix isn't a better chartplotter — the chartplotter is doing its job. The fix is a dedicated NMEA 2000 logger that lives on the bus alongside the MFD, captures every PGN at full rate, buffers offline, and syncs to the cloud when WiFi is available. From there, trip summaries, baselines, and trend detection become straightforward.

That's the category Marine Intel sits in. A small device taps your NMEA 2000 backbone, records every engine and nav message at one-second resolution, and ships the data to a cloud dashboard that learns what your engine's normal looks like. When fuel-rate-at-cruise drifts 6% over eight trips, you get told. The chartplotter keeps doing what it's good at; the logger handles the trend memory it never tried to handle.