Preventive vs Predictive Maintenance

Both approaches live on the same maintenance-schedule object inside AssetAI, which is why plants rarely have to pick one and abandon the other — a pump can carry a time-based PM for lubrication and a condition-based PdM for vibration in parallel, each generating its own work order when its trigger fires.

How the Two Bases Actually Trigger

A time or usage schedule is arithmetic: it holds a meter unit, an interval, and a last/next meter, and it goes due the moment a logged reading crosses the next target — no one has to calculate dates by hand. A condition schedule swaps that math for a monitored parameter and an alarm threshold (vibration in mm/s, temperature, whatever the plant tracks) and issues a PdM work order once a reading breaches it. Both are still rule-based against data someone entered, not a forecast — [VERIFY: no ML or statistical prediction model runs behind this], so the "predictive" label describes trigger logic, not a black-box algorithm.

Where the Data Comes From — and Where It Breaks Down

The accuracy of any condition schedule is bounded entirely by how consistently readings get logged. AssetAI accepts them through the Meter Readings screen, the public QR scan page (with photo OCR filling the value from a gauge photo if the field's left blank), a WhatsApp command, and one other channel — but a threshold rule is only as good as its last entry. A reading that comes in lower than the previous one is rejected outright, since meters and monitored values are expected to move forward only. Common mistakes on the shop floor:

• Treating a condition schedule as "set and forget" — thresholds and intervals stay fixed until someone edits them; nothing auto-tunes from failure history.
• Logging readings sporadically, which turns a PdM schedule into a lagging indicator instead of an early one.
• Confusing RRR scoring with condition monitoring — RRR runs off breakdown frequency, cost-to-purchase ratio, downtime cost and warranty status, and gives a Repair/Review/Replace steer independently of whether any PM or PdM schedule exists at all. It's a complementary signal, not a threshold breach.

Where This Fits in a Broader Maintenance Program

Neither PM nor PdM work orders need a failure cause/remedy on file to close — that gate is reserved for corrective and breakdown maintenance — but every schedule, PM or PdM, still carries a checklist that prints as steps into the generated work order, so the technician always arrives with a task list, not just an alert. For plants formalizing this as policy rather than habit, it's worth reading how these definitions map onto ISO standards and the broader discipline of total productive maintenance, especially for Indian manufacturers scaling up per IBEF's industry data. If you're still deciding whether a CMMS is the right layer underneath either strategy, start with what a CMMS actually is, then look at how condition and calendar-based schedules sit alongside inventory and cost tracking across /features.

Setting Up Both Types Without Overengineering

The practical mistake plants make is treating predictive maintenance as a separate program that needs new tooling before it can start. It doesn't. Because both bases sit on the same schedule object, a maintenance team can stand up a condition schedule the same afternoon it sets up a time-based one — same checklist mechanism, same work-order form, same cost model for parts, services and labour. There's no separate PdM module to license or configure, which matters for plants deciding what to prioritise on a limited pricing plan.

A sensible rollout order:

• Start with time or usage PM on assets where failure modes are well understood — lubrication cycles, filter changes, belt replacements. These need no monitored parameter, just an interval.
• Layer condition schedules onto the same assets only where a measurable parameter actually predicts failure — vibration on rotating equipment, temperature on electrical panels. A condition field with no reliable reading behind it is worse than no schedule at all.
• Use RRR scoring, which runs independently off breakdown frequency and downtime cost, to decide which assets deserve the extra discipline of a condition schedule in the first place, rather than guessing.

India-Specific Considerations

Indian plant floors run on a mix of shift-based manual logging and, increasingly, handheld or shared devices — which is exactly why AssetAI supports four separate ways to log a reading: the Meter Readings screen, the QR scan page with photo OCR for gauges that don't have a digital output, a WhatsApp command for operators who are more comfortable texting a number than opening an app, and one other channel. For plants without dedicated instrumentation — common across mid-sized manufacturing per IBEF's industry data — this means condition-based schedules are achievable using manual gauge readings and a disciplined logging habit, not a capital investment in sensors.

This also sets realistic expectations: a condition schedule here is a threshold rule against whatever gets logged, not a sensor network. Plants evaluating industries coverage or comparing this approach against formal frameworks like TPM should treat AssetAI's PdM as a data-discipline exercise first, and a triggering mechanism second — the rule only fires as often as someone tells it the truth.

Where Checklists and RRR Add Leverage

Every PM or PdM work order — regardless of which schedule spawned it — carries the checklist attached to its schedule, rendered as bullet steps in the work order description. That checklist can also be promoted straight from a Knowledge Card, so a documented fix or inspection routine feeds forward into both preventive and predictive jobs without re-typing it each time. Combined with RRR's standing Repair/Review/Replace signal — which runs off failure count, cost ratio, downtime cost and warranty independent of any schedule — plants get a second opinion on whether a recurring PM or PdM job is even worth continuing, or whether the asset should be flagged for replacement instead. For a fuller picture of how this fits alongside work order lifecycles and cost tracking, see what a CMMS covers or browse all features.