Maintenance strategy should not start with workshops

A maintenance system deployment has a familiar shape. Discovery workshop. Asset criticality workshop. PM strategy workshop, sometimes with a reliability consultant and a criticality matrix on the wall. Template design. Then, twelve to sixteen weeks after kickoff, the first real work order. We have run that programme for eight years and it works. This piece is about why the middle of it is the wrong way round.

What a PM strategy workshop is for

Strip the facilitation away and a PM strategy workshop is a room where people who have read the equipment manuals say what is in them to someone who writes it into a template. The inverter gets an annual visual inspection, a filter clean at the OEM’s interval, a torque check on the DC terminals, thermography on the switchgear. The generator gets a monthly no-load run and an annual load-bank test to the standard. The UPS gets its battery impedance test. Someone knows these things, and the workshop is how the knowledge gets from their head into the system.

The knowledge did not originate in their head. It came from the OEM manual, the warranty conditions, the applicable standard, and the last CMMS export. All of those are documents. The workshop is a reading exercise with a facilitator, and it is expensive because senior people are reading aloud.

Where the workshop goes wrong

Three things happen in the room that do not happen on paper.

Coverage is uneven. The assets people care about get discussed in detail. The met station, the tracker controllers, the fire suppression on the BESS container, the ATS that has not been opened since commissioning, get a line each or nothing. The resulting programme is deep where the room’s attention was and thin everywhere else, and the thin parts are invisible because there is no coverage report for a whiteboard.

Provenance is lost. The interval that goes into the template is whatever was said in the room. Whether it came from the manual, from the warranty schedule, from the previous contractor’s habit, or from a guess, is not recorded. Two years later a new asset owner asks why the inverters are serviced quarterly and the answer is “that’s what we set up”.

The output is a starting point pretending to be a conclusion. Every PM programme is wrong in ways that only the field reveals: fans failing between quarterly services, thirty-six consecutive clean checks on a low-criticality pump. The workshop produces version one and no mechanism for version two, so version one stays live for the life of the contract.

Draft first, then argue

The alternative is to reverse the order. Load the asset register. Attach the manuals and the standards. Produce a draft programme, per asset class, in which every plan carries its source: this task, this interval, from this manual section, or from this standard. List every asset for which no applicable regime was found, explicitly, so the thin parts are visible before anyone is asked to approve anything.

Then hold the workshop. Its job changes from authoring to review. The reliability engineer who used to dictate intervals now reads a draft and disagrees with it in specific places. The disagreements are the valuable part; they are the tacit knowledge the manual does not contain. “The OEM says annual, but on this site the filters are done at six months because of the dust, and here is why.” That correction is worth the room’s time. Reading the manual aloud was not.

Approval happens per plan or in bulk, and nothing enters the live schedule before it. That matters more than it sounds. A draft that can be wrong without consequence is a draft people will read critically.

Version two

The same structure gives the programme a way to change. Every corrective work order is evidence about the preventive programme. Failures between PMs on an asset class are an argument to tighten. A long run of clean inspections on a low-criticality asset is an argument to stretch. Readings that trend across closures are an argument to move from calendar to condition. Each of those can be raised as a proposal, with the work orders that justify it attached, for a planner to approve, adapt, or reject with a reason. Statutory minimums are marked as such and cannot be stretched below the standard, whatever the data says.

That is a maintenance strategy that starts from documents, is corrected by people, and is revised by evidence. The workshop is still in it. It is just no longer where the strategy begins.

What we run today and what we are building

Facterra’s sixteen-week deployment has run in fifty-odd industrial facilities since 2017, workshops included. The draft-first version, where the register and the manuals go in on one side and a cited, reviewable programme comes out the other, is what we are building now, with vertical regime libraries for solar, storage, wind and critical facilities plant as the grounding corpus. A practitioner reviews every regime flagged statutory before it is marked as such.