Equipment Uptime in PCB Manufacturing

Equipment uptime is the share of the planned production time during which a machine is available to produce, and it is one of the few figures in a board shop that translates directly into delivery performance. A machine that is unavailable cannot make boards, and no amount of scheduling can recover the time once it has passed. Improving uptime is therefore less about buying newer equipment than about measuring the losses accurately, analysing the breakdowns that repeat, and correcting the causes rather than the symptoms.

Defining Uptime

Uptime should be defined against planned production time rather than against the calendar, because hours when no work is scheduled are not a loss. The definition also has to decide whether changeover time counts, and the cleanest approach is to report it separately: a machine that is available but changing over is not producing, and the two losses have different remedies even though they both reduce output.

The measurement needs a clear start and end for each stop, which in practice means either automatic recording from the machine controller or a simple operator log. Both work, and the automatic version is better only because it cannot be forgotten. Where a manual log is used, it should be short enough to complete in a few seconds, otherwise the data will arrive at the end of the shift reconstructed from memory.

Unplanned Downtime

Unplanned downtime is the time lost to failures, and its cost is larger than the repair itself. It includes the idle operator, the interrupted batches, the rework caused by stopping mid-process, the interruption to the production flow and the delivery promises that have to be renegotiated. In a shop running at high utilisation, an unplanned stop at the bottleneck also delays every order behind it, which multiplies the effect well beyond the duration of the stop.

Downtime should be recorded in a few categories so that the pattern can be seen: mechanical failure, electrical failure, tooling failure, material or supply problem, and operator intervention. The categories matter because they point at different responses, and a shop that records only the duration learns how much time was lost without learning how to get it back. A mechanical failure suggests a maintenance interval, a material problem suggests a supply issue and repeated operator intervention suggests that the process or the instruction needs to change. Sorting the records by category once a month is enough to show which of them is growing.

Breakdown Analysis

Breakdown analysis starts with the failures that repeat. A machine that fails once a year in a different way each time is a different problem from one that fails monthly on the same component, and only the second is likely to be solved by a change. Sorting the downtime records by machine and by component quickly shows which failures are worth investigating.

The analysis should look for the cause behind the failed part. A contactor that fails repeatedly may be undersized for the load, a bearing that fails early may be misaligned, and an optical unit that drifts may be overheating. Replacing the part restores production and leaves the cause in place, which is why the same repair is performed again a few months later by a different person on a different shift. The record of what was replaced and why is what allows that pattern to be seen at all. The habit of asking why the part failed, rather than only replacing it, is what turns a repair into an improvement.

Mean Time Between Failures

Mean time between failures is the average running time between breakdowns on a given asset, and it converts a history of incidents into a single trend that can be watched. A falling figure means the machine is deteriorating faster than it is being maintained or that the process is being run beyond its design conditions, and either is worth knowing before the delivery performance suffers.

The figure is most useful in comparison rather than in absolute terms. Comparing two drilling machines of the same type, or comparing the same machine this year with last year, shows whether the maintenance effort is working. A single number quoted without a comparison invites arguments about what is normal, which is why the trend matters more than the value, and why the same asset should always be measured in the same way.

Availability and the Bottleneck

Availability work should be concentrated on the constraint, because time recovered anywhere else produces waiting work rather than additional output. A bottleneck machine running three shifts needs both high availability and short changeovers, and the two together determine how much work can pass through the plant in a week.

It is worth calculating what one hour of bottleneck downtime costs, including the contribution margin of the work that could have been produced. That figure converts the maintenance argument from a cost conversation into a value conversation, and it usually justifies maintenance spending that would otherwise be deferred. In most shops the calculation takes ten minutes and changes the discussion permanently, because it converts an engineering preference into a number that a finance director can recognise.

Building a Programme

An uptime programme needs a few elements: accurate recording, a weekly review of the largest losses, a maintenance schedule that is defended, spare parts for the critical components, and a habit of analysing repeat failures. None of these is expensive, and together they move a shop from reacting to managing. The order matters, and recording comes first, because without data every other element is guesswork. A shop that begins by buying spares it may not need will spend money without changing the failure rate.

The programme also needs to survive busy periods, when maintenance is the first thing to be postponed and the first thing whose postponement causes the next crisis. Treating the maintenance window as part of the schedule rather than as an interruption to it is the discipline that makes the programme work, and it is the point at which management support is either present or absent. Everything else in the programme follows from that single decision.

Practical Rules

Define uptime against planned time, record every stop with its category, and review the largest losses weekly. Analyse the failures that repeat, concentrate availability work on the constraint, and keep the critical spares on the shelf.

Record the result with the production records so that the trend can be compared over a year, and connect the improvement to delivery performance rather than presenting it as an engineering metric. Equipment uptime is ultimately a commercial figure, because it decides whether the delivery dates promised to customers are ones the shop can keep.

FAQ

How is uptime calculated? As the share of planned production time during which the machine was available. Changeover time is usually reported separately.

What is MTBF? Mean time between failures, the average running time between breakdowns on an asset. It is most useful as a trend or as a comparison.

Where should availability work start? At the constraint. Recovering time at a non bottleneck machine usually produces waiting work rather than more output.