How Kiosk Punches Feed OEE in EDGEBIC

OEE in EDGEBIC by User Solutions multiplies three factors, availability, performance, and quality, and kiosk punches supply the data behind each one, which is why a work center with no punches shows quality and OEE as n/a. Availability comes from productive hours against shift-derived hours, performance from actual pieces against a theoretical target, and quality from good pieces against all pieces made. Because the quality factor is built purely from kiosk piece counts, honest OEE depends on operators actually punching, which ties this report directly to shop floor data collection.

Overall Equipment Effectiveness is the classic manufacturing scorecard, and its power is that it refuses to let one good factor hide two bad ones. A machine can be available all shift and still score poorly if it ran slow or made scrap. This article shows exactly which punch data feeds each factor and why EDGEBIC would rather show n/a than a fabricated quality number.

The Three Factors

OEE is a product, not an average:

OEE % = Availability % x Performance % x Quality % / 10000

Multiplying is the point. If any factor is weak, the product is weak, so OEE captures the compounding losses a single headline number would mask.

Availability, From Productive Hours

Availability compares the productive hours a work center actually logged against the hours it was available to run:

Available hours = shift hours x calendar days in window x instances
Availability % = min(100, actual productive hours / available hours x 100)

The available hours are shift-derived, not a 24-hour clock, falling back to an eight-hour workday only when no shifts are configured. That matters: a plant running one shift is not punished for the sixteen hours it never intended to run. The actual productive hours come from the kiosk's run and rework punches, which roll up into daily production hours; setup, idle, and down time are excluded from productive hours but recorded separately for downtime analysis. So availability directly reflects how much of the intended shift was genuinely producing.

Performance, From Pieces Against a Target

Performance asks whether the machine ran at rate. EDGEBIC computes a theoretical piece target from the hours actually run and the routing's cycle time:

Theoretical pieces = actual hours x (pieces per unit / cycle time) x efficiency factor
Performance % = min(100, actual pieces / theoretical pieces x 100)

Actual pieces are good pieces from kiosk taps. When a routing has no cycle time to build a target from, performance falls back to the ratio of actual to planned hours. A day that lost time to a quality hold or ran below standard makes fewer pieces per hour, so performance drops, which is exactly the signal you want. For how the piece counts arrive, see how piece counts feed the EDGEBIC schedule.

Quality, From the Piece Counters

Quality is the factor that depends entirely on kiosk punches:

Quality % = min(100, good pieces / (good + scrap + rework) x 100)

Good, scrap, and rework all come straight from the operator's taps, each scrap carrying its reason. This is why the three counters are kept strictly separate, as covered in good, scrap, and rework counts explained. If a work center has no piece punches, there is nothing to divide, so quality is null and displays as n/a. Because OEE multiplies quality in, OEE for that work center is n/a too.

Why n/a Is the Honest Answer

The most important design choice in the OEE report is that a punch-less work center shows n/a rather than a made-up 100%. Fabricating a perfect quality score would flatter every machine that no one measures and quietly reward not collecting data. EDGEBIC refuses to do that. It shows the number only when the punches exist to support it, and it sorts work centers with real OEE first, dropping the n/a rows to the bottom so your coverage gap is visible. The report is honest about what it does and does not know, which is the whole reason it can be trusted for decisions.

A Worked Example

A CNC machine over a reporting window. Operators punch at the kiosk, so all three factors compute.

An OEE of 65.5% reads as a solid day, but the factors point precisely at where the 34.5% went: a fifth of the intended running time was lost to availability. That is a maintenance and scheduling conversation, and OEE surfaced it without anyone guessing. Now imagine the same machine with no kiosk punches: availability and performance might still compute from hours, but quality would read n/a, and so would OEE, telling you plainly that the machine's effectiveness is unmeasured rather than perfect.

What OEE Needs From the Floor

To get a complete OEE row for a work center, operators need to be capturing three things at the kiosk: productive time through run and rework punches, good pieces through the Good counter, and scrap through the Scrap counter with a reason. Configured shifts supply the availability denominator, and a routing cycle time supplies the performance target. Miss the kiosk punches and quality, the one factor only the floor can provide, goes n/a. This is the practical case for the daily punching habit described in the daily rhythm of a kiosk-run shop.

The Bottom Line

OEE multiplies availability, performance, and quality, and each factor is fed by kiosk data: productive hours for availability, good pieces against a routing target for performance, and the good-scrap-rework split for quality. The quality factor exists only where operators punch, so a punch-less work center honestly reads n/a rather than a fabricated perfect score. That honesty is what makes the report worth acting on, and it is one more reason clean shop floor data collection pays for itself. See the platform at EDGEBIC, and how the same data drives production bottleneck identification.