Preventive maintenance replaces parts on a calendar. Predictive maintenance replaces them when the equipment says so. On overhead cranes the difference is measured in unplanned downtime, because a crane almost never fails without warning — it fails without anyone reading the warning.
What a crane tells you before it fails
- Heat where there should not be any — a bearing, a brake dragging, a loose connection on a conductor bar.
- Vibration that changes character. Gearboxes and couplings announce themselves early.
- Wear rate, not wear. A rope losing a wire a month is different from a rope that lost four this month.
- Current draw creeping up on a travel motor: the crane is fighting something, usually the runway.
- Brake travel growing visit over visit — the clearest predictor of a hoist brake that will not hold.
Why the same 66 points, read differently
We use the same inspection structure as the maintenance program, but the value is in the trend, not the visit. One reading is a snapshot; four readings are a direction. That is why the reports are cumulative and why we keep your history — a first inspection is worth a fraction of a fourth one.
The four failures worth predicting, and how each announces itself
Not everything on a crane deserves this attention. Four things do, because each one stops the bay and each one gives warning:
- Hoist brake. The clearest signal in the whole machine. Brake travel grows visit over visit long before the brake slips. Measured and trended, it is close to a countdown.
- Wire rope. Not broken wires, but the rate at which they appear, plus diameter loss at the point that always sits over the sheave. Rope fails predictably if anyone is counting.
- Gearboxes and couplings. Vibration changes character before it changes amplitude. A gearbox that sounds "different" to an operator is usually already three months into a failure.
- Travel motors and wheels. Rising current draw and uneven flange wear both point at the same underlying cause, which is almost never the motor — it is the runway.
Why the runway keeps showing up
More predictive findings trace back to the runway than to any component on the crane. A rail out of level, out of gauge or out of straight makes the machine fight the structure on every pass, and that fight is paid for by wheels, end trucks, motors and eventually the building columns.
This is why a runway survey is part of the program and not an extra. Replacing a crane that "keeps failing" without surveying the rail it runs on is how a plant ends up buying the same failure twice.
What the trend report actually looks like
Each visit produces the 66 points classified as pass, watch or act. What we hand over on top of that is the comparison: which readings moved since last time, in which direction, and how fast. A brake whose travel went from 2.0 to 2.4 to 3.1 mm is a different conversation from three readings of 2.4.
That is the whole method. It is not exotic instrumentation — it is measuring the same things, by the same criteria, often enough that movement becomes visible.
Where it pays and where it does not
Predictive work earns its cost on cranes whose downtime stops production. On a crane used twice a month for maintenance, a solid preventive schedule is the honest recommendation and we will tell you so. We would rather size the program to the risk than sell you the biggest one.
A rough test: if an unplanned failure of that crane would idle people or hold a shipment, it belongs in a predictive program. If it would inconvenience the maintenance department, it probably does not.
Common questions
Do we need to stop production for this?
Most of the inspection is done with the crane available for short windows. Load testing and anything requiring lockout is scheduled with your maintenance lead.
How soon does it show a return?
Usually by the second or third cycle, when the trend data starts flagging things early. The first visit mostly establishes the baseline.
Do you install sensors or monitoring equipment?
Not by default. On most plant cranes, disciplined measurement on a schedule catches the same failures at a fraction of the cost. Where a crane is critical enough to justify continuous monitoring, we will say so — but we would rather earn it with the first three inspections than sell hardware up front.
We already have a maintenance contractor. Does this replace them?
It does not have to. Predictive inspection and repair work are different jobs, and there is an argument for keeping them separate. We are happy to inspect and hand the work orders to whoever does your repairs.