Manufacturing capacity planning is the job of comparing the workload coming in over a period, expressed in hours, with the capacity each work centre can really deliver to find any shortfall or surplus in advance. In practice: convert demand into hours, calculate effective capacity after losses and divide one by the other. Any work centre above 100% is overloaded, and the one with the highest ratio is your bottleneck, so you can decide at the start of the month rather than on the last day.
What happens when sales says "let's take it"
Sales brings in a large order from a new customer and asks whether the due date is achievable. The production manager looks at the machine list and says: "We have three CNCs on two shifts; we'll manage." Two weeks later, older orders, maintenance and changeovers pile up on the same machines. Overtime is booked, one job goes to a subcontractor and a due date is still missed.
The problem is not bad intent; it is that nobody did the sums. Machine count is not capacity. Capacity is the time left after unplanned stoppages, changeovers, maintenance and speed losses have been taken out. Any due date promised without knowing that figure is a guess.
Capacity planning is often confused with production scheduling and APS. Capacity planning answers "will this month's work fit into this factory?" over weeks and months. Scheduling answers "which job runs on which machine, in what order and at what time?" You need to know it fits before you sequence it.
What not knowing your capacity costs
One of the most common mistakes in a capacity calculation is ignoring unplanned downtime:
According to Siemens' "The True Cost of Downtime 2024", large plants average 25 downtime incidents a month and lose 27 hours of production a month. (Siemens, The True Cost of Downtime 2024)
That is a plant-wide average, but for a sense of scale, 27 hours is close to two full working days on a two-shift machine. If downtime is left out, the plan produces the same amount of "work that doesn't fit" every month.
Demand mix also moves quickly. According to OSD's 2025 results, Türkiye's total automotive production rose 4% to 1,419,464 units while car production fell 4%. Total volume can move one way while the mix moves the other. For automotive suppliers, that means one line emptying while another overflows, and only a plan built by work centre shows it ahead of time.
Data infrastructure is part of the problem. According to TurkStat's ICT Usage Survey in Enterprises 2025, only 23.6% of Turkish enterprises with 10–49 employees use ERP software. If routings and operation times are not held in a system, every capacity calculation is done by hand and done incompletely.
The three levels of capacity planning
| Level | Horizon | Question | Inputs |
|---|---|---|---|
| Long-term resource planning | 1–3 years | Do we need a new machine, a new shift or a new site? | Sales forecast, product families |
| Rough-cut capacity planning (RCCP) | 1–6 months | Does the master production schedule fit the critical work centres? | Master schedule, bottleneck work centres |
| Capacity requirements planning (CRP) | 1–8 weeks | Do open work orders fit each work centre week by week? | Work orders, routings, operation times |
For most SMEs the starting point is RCCP: a monthly calculation for just two or three bottleneck work centres reveals most problems in advance.
Manufacturing capacity planning in six steps, with a worked example
- Define work centres. Group machines that do the same work; in our example three CNC machines form one work centre.
- Calculate theoretical capacity. Machines × shifts × net hours per shift × working days. Example: 3 × 2 × 7.5 × 22 = 990 hours.
- Take out the losses. Use real data for an availability rate (planned maintenance, changeovers, unplanned stoppages) and an efficiency rate (speed losses and minor stops). With 85% availability and 90% efficiency, effective capacity is 990 × 0.85 × 0.90 ≈ 757 hours. These rates resemble the availability and performance components you use to calculate OEE; the difference is that a capacity calculation also deducts planned maintenance.
- Convert load into hours. For each order, quantity × unit operation time, plus set-up. Example: product A, 1,200 units × 18 min = 360 hours; product B, 800 × 24 min = 320 hours; product C, 500 × 30 min = 250 hours; set-ups 30 hours. Total load: 960 hours.
- Work out the load-to-capacity ratio. 960 / 757 ≈ 127%. The work centre is overloaded by roughly 203 hours; these orders will not all fit this month under normal working.
- Decide how to close the gap. Weigh the options below for cost and risk, together with the due dates involved.
Building the table across several work centres
In a real factory an order rarely passes through one work centre; it visits cutting, machining, welding and painting in turn. Each operation time on the routing is added to the row of the relevant work centre. You end up with a table of work centres by week, each cell showing that week's load next to its effective capacity. For an example of a part-level flow through panel cutting, edge banding and packing stations, see furniture manufacturing software.
The table also shows the bottleneck moving over time. Welding may overflow in the first week of the month and painting in the last, simply because of routings and the spread of due dates. A weekly view lets you calculate which work centre is relieved by pulling an order forward a week or pushing it back.
Five ways to close a capacity gap
- Overtime or an extra shift. Quick but limited: under Türkiye's Labour Act No. 4857, overtime may not exceed 270 hours per employee per year and requires the employee's consent. Our overtime and timesheet guide explains how it is calculated.
- Shorter changeovers. Running similar products back to back adds capacity without investment; that is the scheduler's job.
- Higher availability. Predictive maintenance cuts unplanned stoppages and enlarges effective capacity directly.
- Moving work to another work centre or a subcontractor. If alternative routings are defined, the load can be balanced.
- Renegotiating the due date. If the calculation is done early, you talk to the customer at the start of the month, not on the last day.
Common mistakes
Planning against theoretical capacity. Loading 950 hours onto a work centre with 990 theoretical hours looks safe on paper. If its effective capacity is 757 hours, that load is roughly 25% over capacity.
Never updating operation times. If routing times were written years ago, the calculation is wrong. A production work order tracking system that collects real times from the floor is the most reliable input to a capacity plan.
Measuring demand only by orders in hand. Make-to-stock businesses need forecast load for the months ahead. AI demand forecasting based on past sales and seasonality extends the RCCP horizon.
Forgetting people. A free machine with no qualified operator on that shift is not capacity. Shift plans and attendance data belong in the calculation. Summer leave, public holidays and machines that only a few people can run all reduce real capacity even when machine hours stay the same.
How we do it at Digital Bridge
We do not sell off-the-shelf packages. A capacity model has to reflect your routings, shift pattern and bottlenecks, so we start with a needs analysis and set out scope, phases and cost in a written proposal.
- Discovery and bottleneck analysis. In a free Industry 4.0 site assessment we map your work centres, routings and critical machines with you.
- Collecting real times. With our MES production management solution, operation times, changeovers and stoppages are recorded on the floor, so availability and efficiency become measurements rather than estimates.
- Linking capacity to the plan. We bring the load and capacity calculation into our production planning and scheduling (APS) solution, where the plan weighs machine capacity, tooling, operator skills and changeover times together, and reports in advance which machine will be overloaded in which week.
- ERP integration. Orders, routings and stock come from your ERP; our MES integrates with Logo, SAP, Mikro and bespoke ERP systems.
- Forecasting and a management view. Where needed, demand forecasting analytics projects load for the coming months, and load by work centre is shown on a BI dashboard.
- Pilot. We start with a monthly RCCP on one or two bottleneck work centres and widen the scope once accuracy has been measured.
For people capacity, attendance data from SmartPass, which calculates timesheets against the shift plan, can feed the model.
How ERP and MES share this work is covered in MES vs ERP, and the role of the ERP itself in our ERP guide for SMEs.
Our Industry 4.0 roadmap for SMEs shows where capacity planning fits in the wider journey; more articles are on the Industry 4.0 topic page.
Your next step
Pick the work centre that causes the most complaints and apply the six steps above to next month in a spreadsheet. If the ratio is above 90%, it is time to turn capacity planning into a regular process. Contact us and we will build the calculation with your own data.