Most preventive maintenance programmes fail quietly. The schedule exists — in a spreadsheet, in a vendor’s diary, in someone’s head — and then a chiller trips on the hottest day of May and the post-mortem reveals the quarterly service was “done” but nobody can show a reading, a photo or a signature.
Here is a practical framework we see working across office parks, hospitals and plants: five checks that tell you whether your PM programme is real, and what to do when it isn’t.
1. Can you list every critical asset with its service interval?
Start with a register, not a schedule. For each building: site → floor → system → component, with make, model, install date, warranty, AMC and the OEM-recommended interval. If the register lives in three places (finance’s fixed-asset list, the vendor’s AMC sheet, the FM’s Excel), reconcile it once and give it a home. QR tags on the asset close the loop — a technician scanning the tag should see history, not a blank form.
2. Does a missed PM create a visible exception?
A schedule that nobody escalates is a calendar, not a control. Each PM should auto-generate a work order with an owner and due date; overdue tasks should escalate to the building manager and show up red on the dashboard the portfolio head opens on Monday. The test: pick a random PM from last month and ask who would have known if it had slipped by two weeks.
3. Is there proof of work on every closure?
Photos, readings (pressures, temperatures, vibration), spares consumed and time on task. Mobile checklists with mandatory photo items make this free for technicians and priceless for audits. For lifts, fire systems and DG sets this is also your statutory evidence.
4. Are you measuring MTBF and MTTR, not just counting tickets?
Ticket counts go up when reporting improves — that is a good thing — so they are a poor measure of reliability. Mean time between failures per asset class tells you whether PM is working; mean time to repair tells you whether your team and vendors are. Trend both by building and compare across the portfolio.
5. Can the data change the schedule?
The endgame is condition-based maintenance: runtime hours and sensor telemetry (vibration, current draw, delta-T) creating work orders before a human notices. That needs an IoT/BMS overlay that reads your existing systems rather than replacing them, and a CMMS that turns alarms into tickets automatically. Even without sensors, meter-based schedules (every 500 DG hours rather than every quarter) beat the calendar.
Where to start
Pick one building and its ten most critical assets. Build the register, switch on PM schedules with proof-of-work checklists, and review MTBF/MTTR after 90 days. Teams that do this typically see asset life extend 2–3× and emergency call-outs fall sharply — and, more importantly, they can show their owners the evidence.
Facyliti’s Preventive Maintenance and Asset Management capabilities implement exactly this workflow — QR-tagged registers, auto-generated work orders, offline mobile checklists and reliability analytics. See it on your own building.
