This article is for plant managers, production managers, production planners and schedulers, and maintenance and continuous improvement teams who regularly see delivery dates confirmed that the shop floor then cannot meet. It explains why a delivery date is only as reliable as the capacity figures used to plan it, and how to replace theoretical capacity in your planning system with the measured availability of your machines.
- Many missed dates start at the handoff between sales and planning.
- Planning systems often use rated capacity, not measured availability.
- Downtime, micro-stops and real changeovers are missing from master data.
- Measured plannable hours make delivery promises more reliable.
What Bernd Vollmüller writes about the handoff between sales and planning
Bernd Vollmüller of Vollmüller & Reese, a German consultancy for production and maintenance, makes a point in his article Die Wahrheit über Produktionsoptimierung: Es geht nicht nur um die Maschinen (The truth about production optimization: it is not only about the machines) that many plant managers know from experience: the largest losses often do not occur at the machine itself, but at the handoffs between departments. He gives a typical interface as an example: “Wenn die Liefertermine des Vertriebs in der Planung landen, hierbei aber u. a. die tatsächlichen Durchlaufzeiten nicht berücksichtigt wurden” (When sales delivery dates land in planning, but among other things the actual throughput times have not been taken into account.)
He describes three levers for a holistic approach: data, process interfaces and clear ownership of the overall result. He also stresses that collecting raw data is not enough: it has to become information you can act on. His article provides the organizational view. This article adds the measurement side: which availability figure for a machine belongs in the plan, where does it come from, and how do you keep it current?
Theoretical capacity and measured availability: two different numbers
In most planning systems, a machine’s capacity is master data: shift time multiplied by rated output, sometimes with a flat allowance for breakdowns that was set years ago. The figure is convenient because it rarely changes. That is exactly the problem. On the shop floor, availability changes all the time.
- Unplanned downtime: breakdowns, missing material, operators waiting. None of it appears in a master data record.
- Micro-stops: stops lasting a few seconds to a few minutes are rarely logged by hand, yet they add up over a shift.
- Changeovers: the standard time in the system rarely matches the real changeover, ramp-up included.
- Speed losses: a machine running below its rated speed reports no stop, but delivers fewer parts.
An illustrative example: planning assumes two full shifts at rated speed for a press. Measurement shows that, on average, the press produces at rated speed for noticeably fewer hours, because micro-stops and long changeovers use up part of every shift. Every order scheduled on that press therefore starts with a delay that only becomes visible once the date is already at risk. Sales confirmed the date correctly according to the plan, production worked correctly, and the delivery is still late.
Where the gap hits the delivery date
The table below shows typical planning assumptions, what automatic measurement reveals, and the consequence for the promised date.
| Planning assumption | What measurement shows | Consequence for the delivery date |
|---|---|---|
| Available time = shift time | Part of the shift is lost to stops and waiting | The order needs more shifts than planned |
| Rated speed for the whole run | The machine runs slower than intended at times | Output per shift falls short of the plan |
| Changeover = standard time | Changeover duration varies with each product switch | With frequent switches, the whole sequence slips |
| Same allowance for every machine | Availability differs widely from machine to machine | The buffer sits in the wrong place |
| Bottleneck according to master data | The machine that really limits output is a different one | Dates are protected on the wrong asset |
Method: bringing measured availability into the plan
The goal is not to reinvent planning, but to give it an honest input. Five steps work well.
- Select the machines that drive delivery dates. Start with the assets where orders queue up or where late deliveries can be traced back to.
- Measure automatically, do not estimate. Over several weeks, capture run time, stops with their causes, micro-stops and real speed, shift by shift.
- Derive plannable hours. For each machine and shift pattern, calculate how many hours at rated speed you actually get, and look at the spread as well: a machine that varies a lot needs more buffer than one that runs steadily.
- Replace the master data. Enter the measured values instead of theoretical capacity in your ERP or planning system, and place buffers deliberately in front of the real bottleneck.
- Reconcile regularly. Compare plan and actual every week and update the values after improvement projects, product mix changes or major maintenance work.
The fifth step matters most. A value measured once goes stale just like an estimated one. Only ongoing reconciliation turns measurement data into the actionable information Bernd Vollmüller describes.
Who contributes what at the interface
Measured availability is a shared number, not a production number alone. For it to work at the interface, each function needs a clear role.
- Sales: confirms dates based on the capacity that planning derives from measured values, not on rated output.
- Production planning: maintains the plannable hours per machine and flags when the load on the bottleneck exceeds its measured availability.
- Production and maintenance: tackle the biggest causes of downtime at the bottleneck, because every hour recovered there is extra delivery capacity.
- Plant management: owns the overall result, meaning the date that is kept, not the utilization of individual departments.
TEEPTRAK provides the measurement layer for this. A module is wired into the electrical cabinet and a sensor is mounted at the edge of the line; installation takes about one hour per machine and needs no connection to the PLC. OEE, downtime and micro-stops are available within 48 hours, and operators assign stop causes directly at the machine. TEEPTRAK does not replace your ERP or planning system: it feeds them measured availability. The solution is used in more than 700 plants in over 30 countries.
Further reading
- Die Wahrheit über Produktionsoptimierung: Es geht nicht nur um die Maschinen (Bernd Vollmüller, Vollmüller & Reese, in German)
- Presenting production losses to management: from percentages to hours and capacity
- Realistic OEE target: set it machine by machine, in steps
FAQ
Why is rated capacity not enough for scheduling delivery dates? Because it leaves out downtime, micro-stops, real changeover times and speed losses. A date built on rated capacity relies on hours the machine does not deliver in practice.
How long should you measure before changing planning values? Long enough to capture normal variation: several weeks that include every shift pattern and the main product changeovers. After that, update the values on a regular basis.
Do you need to replace your ERP system? No. The measured plannable hours replace theoretical capacity in the master data of your existing system. What matters is that someone owns these values and reconciles them with actual output regularly.
0 Comments