botminds.aiRequest a demo
WorkflowAsset Management, Maintenance & Reliability

Preventive Maintenance Optimization

Preventive maintenance optimization that shows what each routine has really found, why it was written, and who signed the change.

See the demoBook a demo
Follow 612 motors from the plot to a signed change ↓
The problem

Tens of thousands of routines, and no one can say which of them find anything

A routine is copied from a maker’s manual into the first maintenance plan, moves through two system changes, and arrives today with its basis written as OEM. The motors it was written for left the platforms long ago. The crews still do it four times a year, and the closeouts say done.

Reading a field’s closeouts by hand would take years of someone’s time, so a study team samples one class at a time. Each change needs a change record, and nobody starts one without proof. Reviews that hunt for hours rarely look for the failures landing between visits.

Built forMaintenance strategy engineersTechnical authoritiesMaintenance plannersField maintenance leadsInternal auditors

The Problem Today

Routines outlive the equipment they were written for, and nobody can show which ones ever find anything.

98,500
hours to read one field’s closeouts at 30 seconds each our estimate

Reviewed by hand today

  • Routines
  • Closeouts
  • Manuals
  • Intervals
  • The reason is lostSystem moves dropped why each task exists
  • Failures between visitsSavings reviews rarely look for what gets missed

The Solution

  1. Read each closeoutEach note called: a real finding, no finding, or nothing written
  2. Trace why it existsEach routine followed back to its manual, change record or failure
  3. Bundle by equipmentKeep, change or tighten; hidden failures judged on test results
  4. Two people signAn engineer decides; the technical authority opens a sample and signs

Built-in controls

  • Safety and legal tasks locked; doubts held
  • Riskiest routines opened before signing
  • Each change watched a year on

What You Get

Crew hours go to routines that find things, and every interval change carries its reason and its signature.

  • Hours put backRoutines that find little relaxed, net of what tightening adds
  • Gaps closedTasks that miss failures between visits tightened
  • Changes you can defendEach one with its source, its sample and a signer
One bundle, start to signature

Preventive maintenance optimization on 612 motors inspected four times a year

An offshore field with three platforms carries 31,400 routines and 214,000 booked hours a year. Here the strategy engineer takes the quarterly inspection on its low-voltage motors to a signed change.

demoThe field, its platforms, the people and the hours are invented for the sample.

  1. 01Morning

    Each routine on one plot: hours a year against what it really finds

    Maintenance strategy engineer

    Agents have read 1,182,640 closeouts and placed every routine by its hours and its real findings per 1,000 visits. Top left costs hours and finds almost nothing: 6,588 routines, of which 3,318 carry a proposal and 2,630 are locked. Bottom right finds a fault so often that failures may be slipping between visits: 85 routines.

  2. 02Morning

    612 motors, 5,800 hours a year, and 3 real findings

    Maintenance strategy engineer

    She picks the cluster of low-voltage motors, the bundle the agents put first for her: a quarterly inspection costing 5,800 hours a year, with 3 real findings in 19,584 visits. The rest of the field fades so the one bundle stands out, and the card beside it reads relax to yearly, 4,350 hours a year off, waiting on her decision.

  3. 03Closeouts

    19,584 closeouts read, and the three that found something

    Maintenance strategy engineer

    Each closeout is called by what it says. 17,902 are clean, 1,611 say only done and count as unknown, 68 look like findings but are not, and 3 are real: a perished gasket with water in the box, a cracked fan cover, a loose cable gland. She can open each note and read it in the technician’s words.

  4. 04History

    Written for open bearings, copied into the first plan, then labeled OEM

    Maintenance strategy engineer

    The trail runs six links. A maker’s manual asked for quarterly care of open bearings; the first maintenance plan copied it word for word; a system move dropped the basis; the motors were replaced by sealed-bearing units; a second move wrote OEM. The fitted motors’ own manual says yearly. Each link opens its source.

  5. 05The case

    186 corrective jobs in 8 years, and the inspection saw 3 of them coming

    Maintenance strategy engineer

    The case puts the routine beside the failures. Of 186 corrective jobs on these motors, the vibration route or an operator found 141 and 42 gave no warning. A motor fault shows itself by tripping or by noise, so nothing here is hidden. It also lists what still protects, the 90 locked safety-critical motors and the hours on both sides.

  6. 06Decision

    Yearly, a reason given, and the change record drafted

    Maintenance strategy engineer

    She picks yearly and a reason: the original reason no longer applies. The expected corrective work a longer interval may add is on the case, 17 hours a year, beside the 4,350 coming off. Her decision drafts the change record, and the next step is to send it to the technical authority. She cannot sign it herself.

  7. 07Sign-off

    Five sampled motors, picked by past findings and the harshest service

    Maintenance technical authority

    The technical authority’s check opens with five routines, the riskiest in the bundle: picked by past findings, then harshest service, then longest run. The first is a seawater booster pump motor that had the terminal box finding. He reads its last visits and marks each sampled routine fine to include. Nothing signs until all five are opened.

  8. 08Sign-off

    The change record says what still protects and when it is checked again

    Maintenance technical authority

    Before signing he opens the change record. It sets out what changes, why, what still protects, what is left out and locked, the risk before and after, and the maker check. It also says that the old interval is held on the record, and that the change is checked again 24 months after loading.

  9. 09Signed

    Signed, with 5 of 5 sampled routines opened

    Maintenance technical authority

    With the five opened and the change record read, he signs. The bundle reads signed and ready to load, and the planner takes it from there. Sample size for this check was set when it opened; a later change in settings does not touch it.

  10. 10After

    Signed so far: 7,557 hours, out of 17,739 net if all are signed

    Maintenance strategy engineer

    Back on the plot, the motors’ proposal reads signed. The footer keeps the running sum honest: 19,100 preventive hours off, 1,282 on from tightening and 79 expected corrective hours added, for 17,739 net if everything proposed is signed. Lost production and failures avoided are not counted.

17,739 net hours, after tightening and the corrective work it may add

If every bundle on the field is signed as proposed, a year looks like this.

Preventive hours taken off19,100
Hours added by tightening−1,282
Expected corrective hours added−79
Net hours a year17,739

demoSample data, from the working solution.

PM interval optimization, bundle by bundle, in both directions

Some intervals go out and some come in. Each row is one bundle on the field.

BundleNow, then proposedReal findingsHours a year
Low-voltage motorsQuarterly, then yearly3 in 19,584−4,350 h
General area lightingMonthly, then yearly6 in 20,352−3,498 h
Ventilation fan belts and filtersWeekly by two people, then one10 in 51,584−3,224 h
Pressure transmitters, clean serviceYearly, then every 3 years7 in 10,192−2,718 h
Seal gas filtersYearly, then every 6 months96 in 328+328 h
Cooling seawater strainersQuarterly, then monthly in summer74 in 832+234 h
Export pump seal barrier potsMonthly, then weekly211 in 1,728+720 h

demoSample data, from the working solution. Valves and small pumps fail silently, so their bundles are judged on test results rather than finding counts.

Around the motors

Tightening, field checks and the look-back behind each PM task review

Five screens the strategy engineer and the field lead work from, and the one an auditor reads.

Bundles, largest first11 proposed, 3,467 routines, each with its findings, hours and where it stands: signed, with the technical authority, or waiting on a field check.
Tasks that need tighteningSeal gas filters: 96 of 328 changes found a fault, and 7 seal events fell between changes. Proposed every 6 months, with the pressure rule back in.
Questions only the crew can answerDo the sump pumps stick if they sit a quarter without running? The field lead answers; the strategy engineer settles.
One year laterChanges signed a year ago, faults before and after. Gearbox oil level went from 2 faults to 7, so the agent suggests going back. A person decides.
A chained audit trailEach decision, opened sample and signature on its own line, each chained to the one before, with an evidence pack holding every signed change record.
Rules for the checkSampled routines per check, days to sign, the share of done-only closeouts that stops a proposal, and the look-back. The engineer proposes; the technical authority approves.

demoSample data, from the working solution.

Three things the closeouts and the history showed that a sample would miss

A task written for motors that are gone

The quarterly inspection was for open bearings. Sealed-bearing motors replaced them, and the basis was lost in a system move.

A rule that fell out in a migration

The maker said change seal gas filters at a pressure limit or yearly, whichever came first. Only the year survived, and seal events fell between changes.

A change going the wrong way

A year after the gearbox oil level check was relaxed, faults rose from 2 to 7, so the agent proposes going back.

demoSample data, from the working solution.

Behind the plot

Agents that read every closeout, and none can relax a locked routine

Each one proposes or reports. People decide.

Closeout reader

Reads each closeout and says whether it found something real.

  • Never counts done as clean
  • Below 80% sure, asks the field lead
Task historian

Traces each routine to the manual, change record or report behind it.

  • Never invents a reason
Lock keeper

Checks the safety-critical register and legal lists.

  • Never removes a lock
  • Keeps a doubtful match locked
Bundle builder

Groups routines by class and failure and proposes keep, change or stop.

  • Never proposes relaxing a locked routine
Gap finder

Finds failures between visits and repeat findings.

  • Never tightens without showing the failures
Change writer

Drafts the change record and the load file.

  • Never signs
  • Never writes to the maintenance system
Look-back watcher

Compares faults before and after each change.

  • Never reverses a change

demoSample data, from the working solution.

What strategy engineers ask about preventive maintenance optimization

What is preventive maintenance optimization?

Preventive maintenance optimization is the review of routines and intervals against what they really find. Here agents read each closeout, trace why each routine exists, group routines into bundles and propose keep, relax, stop or tighten. The strategy engineer decides and gives a reason; the technical authority opens a sample and signs.

How is PM optimization done safely on safety-critical equipment?

Safety-critical and legally required routines are locked against relaxing, and a doubtful match stays locked. On a standby safety device, finding nothing may just mean the device works, so those are judged on test results.

Does a maintenance strategy review only cut hours?

No. The same reading looks for failures between visits and repeat findings, and proposes tightening where they show up. The net figure takes off what tightening adds and the corrective work a longer interval may bring.

How does it reduce the planned maintenance burden offshore?

By taking hours off routines that find almost nothing, net of tightening, so crew time goes where it finds faults. The hours and the beds they free are shown, and lost production is not counted.

What happens after a change is loaded?

A look-back compares faults before and after each change, waiting at least two gaps of the new interval. If faults rise, the agent proposes going back and a person decides.

What if the closeouts just say done?

A closeout that only says done counts as unknown, never as clean. A routine where too many closeouts say only done is held as not enough evidence and gets no proposal.

Asset Management, Maintenance & Reliability

Other reliability and maintenance planning solutions

All 10 solutions →
Maintenance Technician Assistant

The right page at the machine, and a spoken closeout a supervisor can sign.

1-min demoAsset Register Management

Every finance line traced to maintenance and the plant, with proof.

1-min demoAsset Hierarchy Management

A flat list becomes a working tree, and every move shows what follows before it lands.

1-min demo

Bring one class of routines
nobody has dared to change.

We’ll walk you through its closeouts, its history and a case your technical authority can sign, using your own records.

Book a demoBrowse solutions