Production Capacity Planner

What this tool does

CM8-03 is a weekly rough-cut capacity planner. You describe your resources (machines or work centres with a shift pattern and realistic losses) and your demand (products with weekly quantities and routing minutes per unit). The tool computes the load on every resource, flags bottlenecks, shows how much of the demand mix actually fits, and lets you test fixes — an extra shift, another machine, better efficiency — without touching the baseline plan.

Everything runs inside this single HTML file. No installation, no account, no network requests — your shift patterns, routings and demand figures never leave your computer.

The capacity model

All figures are per week. For each resource:

For each resource, the load placed on it by all products:

Planned downtime and OEE

Two separate deductions keep the model honest:

• Planned downtime % — scheduled non-production time: preventive maintenance, cleaning, breaks that stop the machine, planned changeover windows. Typical planning values: 2–10%.
• OEE / efficiency % — how much of the remaining time produces good parts at rated speed. It bundles unplanned stops (availability), slow running (performance) and scrap/rework (quality). World-class discrete manufacturing is ~85%; typical plants run 55–75%. If you don't measure OEE, start with a conservative 70–80% and refine.

Routings and setup allowance

• Minutes per unit is the standard run time (cycle time) of one unit on that resource. Leave it 0 where a product doesn't visit a resource.
• Setup allowance % spreads changeover time over the parts produced. If a 60-minute setup precedes a typical batch of 100 parts at 6 min/part (600 min run), that is 10%. Small batches push this figure up fast — one of the quickest capacity wins is larger batches or faster changeovers (SMED).
• A product's standard hours column shows its total weekly workload across all resources — useful for sanity-checking routings against your ERP standards.

Reading the results

• OK load ≤ 85% — normally plannable; slack absorbs variation.
• Near limit load > 85% — queues grow quickly at high utilisation; expect longer lead times and little tolerance for breakdowns or rush orders.
• Bottleneck load > 100% — the demand mathematically does not fit. Something gives: overtime, subcontracting, late orders or lost sales.

Feasible mix answers “how much of this week's demand actually fits?” It scales the whole product mix down proportionally until the worst-loaded resource hits exactly 100%. A feasible mix of 93% means you can ship 93% of every product's demand; the per-product table shows the resulting quantities. Over 100% means the whole mix fits with headroom. (Changing the mix — prioritising products with less bottleneck content — can beat proportional scaling; that is classic Theory-of-Constraints territory.)

The bar chart plots load against a common axis with dashed markers at 85% and 100%. The tab badge counts flagged resources and turns red when any resource exceeds 100%.

What-if scenarios

The what-if panel is a non-destructive overlay per resource:

• Δ Machines — add (or remove, with negatives) identical machines.
• Δ Shifts/day — add or drop shifts of the same length.
• Δ OEE points — model an improvement project, e.g. +5 points from faster changeovers.

While a scenario is active, the load table gains scenario columns, each bar shows the baseline as a thin bar beneath the scenario bar, and the what-if table shows the effect in percentage points. “Clear what-if” returns to baseline; your inputs are never modified.

Accuracy and limits

This is rough-cut capacity planning (RCCP) — the standard sanity check between a demand plan and detailed scheduling. Deliberate simplifications:

• Weekly buckets, infinite queues: hours are compared in weekly totals; sequencing, batching interactions and queue timing within the week are not modelled.
• Homogeneous machines: all machines in a resource are interchangeable and share one shift pattern. Model a faster/slower machine as its own resource.
• Deterministic standards: demand and cycle times are single numbers. At loads above ~85%, real-world variability makes performance markedly worse than the average suggests.
• Labour not modelled separately: if operators, not machines, are the constraint, model the crew as the resource.

Use the result to make capacity decisions (shifts, machines, subcontracting, demand levelling) — then verify with detailed scheduling before promising dates.

Saving your work

• Autosave: the whole plan is kept in this browser's local storage automatically (see the indicator in the top bar) and restored next time you open the file.
• Save: writes to browser storage straight away and confirms with “Saved ✓” and the time — useful before closing the tab. It is not a file: use Export .json for that.
• Report details: title, company, prepared-by and closing notes (under the load table) appear on the printed report and are saved with the plan.
• Save project (.json): saves resources, products, routings and the active what-if scenario to a file you can archive, share or reload.
• Export CSV: the resource-load table (including scenario columns when active) for spreadsheets and S&OP packs.
• Print report: a clean report — header, summary tiles, the three executive charts, the load table, demand & routings, any active what-if and your closing notes. It is rebuilt from the current data every time you print (browser print dialog → save as PDF works too).