22/02/2026
Predictive Maintenance, PowerPoint Economics, and the Myth of “Operational Efficiency” - Or "why you should keep accountants in their box"...
Predictive Maintenance (PdM) is sold as engineering progress. In reality, it’s often an accounting fantasy wrapped in sensors and dashboards.
I’ve seen PdM rolled out on single yachts and fleets. The pitch is always the same: fewer failures, lower costs, smarter operations. The outcome is also usually the same—engineers spending more time maintaining sensors, wiring looms, and false alarms than the actual machinery.
Here’s the part that rarely makes it into the presentation:
In real marine conditions—salt, vibration, humidity—sensors fail far more often than engines. Thousands of channels mean constant recalibration, cable faults, and bad data. Scale doesn’t magically fix this. In fleets, you just multiply the problem.
At some point, PdM stops being predictive maintenance and becomes sensor maintenance at scale.
Do the sums; Average of 3000 sensors, mean time between failure of sensor and associated wiring 5 years(?). That makes an astonishing 11 failures a week.....
Why does it keep getting approved?
Because it looks phenomenal in a spreadsheet.
Accountants love PdM:
clean ROI models
tidy MTBF assumptions
dashboards that imply control
But operations don’t live in PowerPoint. They live in heat, noise, corrosion, fatigue—and people making judgment calls at 02:00 to keep assets running.
A good engineer on rounds will hear bearing noise, smell insulation breakdown, and see corrosion long before an algorithm has “enough data.” Add disciplined inspections and a simple log or spreadsheet with manufacturers mean time between failure data, count the hours, and you get context-aware maintenance with almost zero failure overhead.
Accountants aren’t the enemy—but they should not be driving operational design. When finance leads ops, systems get optimized for reporting, not reliability. Complexity increases, resilience drops, and the risk gets pushed onto the people doing the real work.
The best setups I’ve seen are simple:
Engineering defines reality
Operations sets priorities
Finance supports decisions after the physics are understood
PdM has its place—but until sensors are much more reliable than the machines they monitor, most programs protect the data more than the asset.
If your maintenance strategy only works on a slide deck, it won’t survive first contact with reality.
The image is myself in the Control Room of one of the first diesel-electric Azipod yachts to be delivered 20 years ago with project management from Central Yacht.
Curious how others have experienced PdM in the wild—helpful tool, or beautifully packaged overhead?
By, Captain Paul Brackley, Central Yacht, CEO