The production manager knows that the press line was down for two hours on Tuesday because of a tooling change. The maintenance bloke knows it was actually closer to three because they had to wait for a part. The job card says seven units came off, but two were reworked. The ERP says eight were completed because someone entered the rework as a new job. And the monthly report that lands on the owner's desk says "on target" because it only counts invoiced jobs.
That gap between what happened on the floor and what the numbers say is where most small manufacturers lose visibility. OEE — overall equipment effectiveness — is the standard way to measure it. The concept is straightforward. Getting the data is the hard part.
What OEE is, in plain English
OEE measures how much of your planned production time is actually productive. It combines three things.
Availability is the share of planned time the machine was actually running. If you planned an eight-hour shift and the machine ran for six hours after changeovers, breakdowns and waiting for material, availability is 75 per cent.
Performance is whether the machine ran at its intended speed while it was running. If the target is 10 units per hour and it actually produced 8 per hour, performance is 80 per cent.
Quality is the share of output that was right first time. If 48 units came off the line and 45 were good, quality is about 94 per cent.
OEE is the three multiplied together. In this example: 0.75 x 0.80 x 0.94 = 0.564, or about 56 per cent. That means just over half the planned production time produced good output.
A worked example (clearly fictional numbers)
Suppose a small fabrication shop runs a CNC machining centre for one shift, five days a week.
- Planned production time: 8 hours (480 minutes)
- Downtime: tooling changeover 45 minutes, waiting for material 30 minutes, unplanned breakdown 15 minutes. Total downtime: 90 minutes.
- Actual running time: 390 minutes. Availability: 390 / 480 = 81 per cent.
- Target cycle time: 6 minutes per part (10 parts per hour).
- Actual output during running time: 55 parts. Ideal output at full speed would be 65 parts (390 / 6). Performance: 55 / 65 = 85 per cent.
- Good parts: 52 out of 55. Quality: 52 / 55 = 95 per cent.
- OEE: 0.81 x 0.85 x 0.95 = 65 per cent.
That 65 per cent is not unusual for a small manufacturer without automated data collection. The value of OEE is not the number itself — it is what it shows you. In this example, 19 per cent of planned time was lost to downtime (the availability gap), 15 per cent of running capacity was lost to slow cycles (the performance gap) and 5 per cent of output was lost to quality (the quality gap). If you only had one place to improve, you would start with the downtime.
Why the data is hard to get
For a large manufacturer with a manufacturing execution system (MES) on the shop floor, the data comes from sensors, machine interfaces and barcode scans. For a small manufacturer with five to fifty staff, the data usually comes from people writing things down.
Machine run time is often estimated or recorded at the end of a shift on a paper log or a whiteboard. Start and stop times might be rounded to the nearest half hour. Changeover time, which is the gap between one job finishing and the next one starting, is frequently not recorded at all.
Downtime reasons are where the picture gets blurry. Planned downtime (changeovers, maintenance) and unplanned downtime (breakdowns, waiting for material, waiting for instructions) need different treatment in OEE, but the log on the whiteboard just says "down 2 hrs". Without the reason, you know the number but not where to fix it.
Output counts usually come from the job card or the production order. But if rework is entered as a separate job, or if scrap is not recorded until stocktake, the quality figure is wrong. And if the output is entered into the ERP at the end of the week in a batch, the daily picture is missing.
Cycle times — the target speed for each product on each machine — often live in the production manager's head or in a spreadsheet that has not been updated since the machine was set up. If the target is wrong, the performance number is meaningless.
The result is that most small manufacturers either do not track OEE at all, or they track it monthly in a spreadsheet with numbers that are too rough to act on. The monthly average smooths out the day-to-day problems, which is exactly where the improvements are.
What a simple production report looks like
You do not need a full MES or a shop-floor system to start measuring production. A simple daily production report that the shift supervisor fills in at the end of each shift can capture enough to calculate OEE and spot the patterns.
The report needs five things for each machine or work centre, for each shift:
- Planned production time. How many hours the machine was supposed to run.
- Downtime events. Each period the machine was not running, with a start time, an end time and a reason (changeover, breakdown, waiting for material, waiting for tooling, no operator, planned maintenance).
- Output. How many units or jobs were completed.
- Rejects or rework. How many were not right first time.
- Product and job reference. What was being made, so you can compare the actual cycle time against the target.
That can be a paper form, a spreadsheet on a tablet at the machine, or a simple digital form that feeds into a shared spreadsheet or database. The format matters less than the discipline: if it is filled in at the end of each shift, the data is useful. If it is filled in from memory on Friday afternoon, it is not.
Once you have daily data, even in a spreadsheet, you can calculate OEE by machine, by shift, by product and by day. The patterns will show up within a few weeks: the machine that always loses time to changeovers, the shift that records more rework, the product that takes twice the expected cycle time.
Connecting production data to cost data
OEE tells you about production efficiency. But what the owner and the finance manager usually want to know is what a job cost. That means joining the production data (hours, output, downtime) to the cost data (materials, labour, overheads) from the accounting file or ERP.
In most small manufacturers, this join happens in a spreadsheet at month end. Someone exports job hours from the timesheets or job cards, material costs from purchase orders or stock issues, and overhead allocations from the accounting file, and tries to put them together.
The problem is that the data rarely lines up cleanly. The timesheet records clock-on and clock-off, not which job was being run. The material issue was entered as "steel flat bar" with no job number. The overhead is allocated by a formula that has not changed in three years. By the time someone has reconciled it all, the month is over and the numbers are six weeks old.
What a sensible production reporting setup does is connect the production record (job, machine, hours, output, scrap) to the cost records (materials issued, labour hours by job, overhead rates) so the cost per unit or per job is available weekly, not monthly. It does not replace the accounting file. It sits alongside it and draws on the same data.
The production data integration service is about making that connection: getting production records, timesheets, material issues and the accounting file to talk to each other so the production report and the financial report agree.
MRP: when searches rise, the question is the same
Searches for MRP systems and manufacturing resource planning are climbing in Australia. The question behind most of them is: "Can we plan production without the spreadsheet?" If the fabrication shop is scheduling jobs on a whiteboard and checking material availability by walking to the stores, that is a sign the spreadsheet or the manual process has hit its limit.
MRP does not replace the need for good production data. It depends on it. An MRP system needs accurate BOMs (bills of materials), lead times, stock levels and job schedules to plan production. If the stock in the system does not match the shelf, the plan is wrong from the start.
Before buying an MRP, it is worth asking whether the data is ready. If the BOM is in a spreadsheet, stock levels are only right on stocktake day and job schedules live on a whiteboard, the first step is getting those basics into a system you can trust. The production planning post walks through the order to move things off Excel in a fabrication shop.
When you do not need OEE
Not every small manufacturer needs to track OEE. If you run a job shop where every job is different, the cycle time target changes every day and the meaningful measure is job cost versus quote, not machine efficiency. OEE is most useful when you have repetitive production on the same equipment: the same machine making similar parts day after day, where small improvements in uptime or speed compound over thousands of units.
If your bottleneck is not the machine but the materials (waiting for supplier deliveries), or the paperwork (waiting for drawings or instructions), or the people (not enough trained operators), OEE will show you that, but the fix is not on the shop floor.
And if you are a very small operation — one or two machines, the owner is also the operator — a full OEE tracking regime is probably more overhead than it is worth. Keep a simple log of downtime and scrap, know your job costs, and invest the time in quoting and production rather than measurement.
Frequently asked questions
What is OEE in manufacturing?
OEE stands for overall equipment effectiveness. It measures how much of your planned production time produces good output. It combines availability (was the machine running?), performance (was it running at speed?) and quality (was the output right first time?). It is a single number, expressed as a percentage, that shows the gap between what a machine could produce and what it actually produced. It does not account for demand — if the machine was not scheduled to run, that time is excluded.
What is a good OEE score for a small manufacturer?
There is no universal benchmark. A score of 85 per cent is sometimes quoted as "world class", but that figure comes from large, highly automated production lines. For a small Australian manufacturer with manual data collection and frequent changeovers, 50 to 70 per cent is common. The important thing is the trend and the breakdown: whether you are improving, and whether the losses are in availability, performance or quality.
How do I calculate OEE without sensors or a shop-floor system?
You can calculate OEE from a simple daily production log. Record planned production time, downtime events with reasons, output count and rejects for each machine and shift. From those four inputs you can calculate availability, performance and quality, and multiply them for OEE. A spreadsheet is enough to start. The discipline of filling in the log consistently matters more than the technology.
What is the difference between OEE and production reporting?
OEE is one measure within a production report. A full production report might also include job cost versus quote, on-time delivery, WIP value, scrap rates by product and labour utilisation. OEE focuses specifically on machine or line efficiency. If you are a job shop where every job is different, the cost-per-job report may be more useful than OEE.
Does OEE work for a job shop?
It can, but it is less meaningful when every job has a different cycle time and setup. OEE is most useful when the same machine produces similar parts repeatedly. For a job shop, tracking downtime, changeover time and scrap is still valuable, but the primary measure is usually job cost versus the quoted price. The estimate versus actual post covers how to compare finished jobs in construction and manufacturing.
If you want to start tracking production data and you are not sure where to begin, get in touch for a free initial chat. We will tell you honestly whether a system will help or whether a simple spreadsheet and some discipline will do.