Factory Automation ROI: A Worksheet for Downtime, Scrap and Ongoing Costs
Build an automation business case that separates cash savings from capacity, tests downtime and scrap assumptions, and includes ongoing costs.
Use the interactive tool ↓
The same lost hour can appear twice
A proposed packing cell saves operator time and recovers production lost to minor stops. Both improvements may be useful. They do not automatically create two independent cash savings. If the operator remains on the same paid shift, the first benefit is available time. If the recovered output cannot be sold or used to avoid another cost, the second is available capacity. A worksheet should preserve those distinctions before anyone calculates payback.
Consider a hypothetical manufacturer deciding whether to automate a recurring packing constraint. Sales has an order forecast, production has a downtime log, and finance has a margin estimate. Each looks plausible on its own. The investment becomes defensible only after the team checks that they refer to the same products, periods and achievable operating conditions. Start there, before comparing the machinery quotations.
Excellence industrielle’s 2025 annual report describes manufacturers delaying projects under economic uncertainty and a greater emphasis on prioritization and risk reduction. It does not establish a payback target for your factory. That uncertainty is a reason to make assumptions inspectable, rather than hide them inside a single optimistic return percentage. [1]
Build the baseline around the decision
Choose a representative period that includes normal product variation, changeovers and ordinary interruptions. Record scheduled time, relevant downtime, good output, rejects and rework. Keep the original reason codes alongside any cleaned categories. A machine alarm may mark when a line stopped without explaining whether material, maintenance or a quality hold caused it.
Compare the production record with order demand and the observed constraint. Recovering an hour on a machine that normally waits for the next process may produce no extra saleable output. Recovering an hour on the constrained line matters only when upstream supply, downstream capacity and demand can use it. Have planning and operators review the proposed conversion from recovered hours to additional good units.
Name the person responsible for each assumption and record its evidence date. Use actual scrap disposal and rework costs where available, not the selling price of every rejected part. Mark estimates as estimates. If the baseline is incomplete, a short measurement exercise can be the next investment instead of treating missing data as zero loss.
Separate three kinds of benefit
The first category is avoidable cash cost. Examples include overtime that will genuinely stop, purchased rework that will no longer be needed, or material consumption reduced at the same accepted output. Finance should confirm which spending changes, when it changes and whether a minimum purchase or staffing commitment prevents the saving.
The second is additional contribution from output that can actually be sold. Multiply incremental good units by contribution per unit after the variable costs of producing and selling them. Do not use revenue as profit. Cap the units by credible demand and by the remaining production constraints. If the project simply moves output from one week to another, explain the timing value separately.
The third is capacity, resilience or management time that has value without a verified cash conversion. Show these beside the financial result. A supervisor spending less time reconstructing stoppages can be an important reason to invest, but it should not become a salary saving unless an actual budget change follows. Avoid counting the same recovered units once as downtime benefit and again as reduced scrap.
Put installation and continuing work into the model
One-time cost includes equipment, tooling, interfaces, engineering, guarding and qualified safety work, freight, commissioning, training and the planned production interruption. Ask for exclusions in writing. Include site preparation and internal project effort where they consume budget or scarce capacity, and identify which estimates remain unquoted.
Annual cost includes software, support, maintenance labour, spare parts, calibration where needed, connectivity, energy changes and recurring review of data definitions. Budget for the people who investigate exceptions. A connected cell that nobody maintains can keep producing numbers after those numbers stop describing production.
Keep currency, tax treatment and timing consistent with finance’s preferred investment method. This simple worksheet is a screening tool, not a replacement for a discounted cash-flow or financing assessment. Put uncertain expenditure in a visible contingency line with a reason rather than burying it in an apparently exact equipment price.
Work through a deliberately small example
The following numbers are illustrative, not a Quindart result or an industry benchmark. Suppose total one-time cost is C$120,000. The team expects 100 recovered constraint-hours per year and 20 additional good units per hour. After checking demand and other constraints, it accepts 2,000 incremental units at C$12 contribution each: C$24,000 of annual contribution.
Separately, the project avoids C$10,000 of material and disposal costs at the original output level, and C$8,000 of overtime that finance confirms can be removed. Those benefits do not overlap the incremental-unit calculation. Annual support, maintenance and other additional operating costs total C$12,000. Annual net cash benefit is therefore C$24,000 + C$10,000 + C$8,000 − C$12,000 = C$30,000.
Simple payback is C$120,000 ÷ C$30,000 = four years, assuming that annual benefit starts immediately and remains constant. That assumption is often too neat. If commissioning delays benefits or demand falls, the timing changes. If only half the incremental units can be sold, net annual benefit becomes C$18,000 and simple payback becomes about 6.7 years. A non-positive net benefit has no positive simple payback; do not display it as zero years.
Use the worksheet to decide what must be proven
Complete conservative, base and upside cases using different recoverable downtime, sellable units, scrap improvement, ramp-up and support cost assumptions. Change the uncertain drivers, not just the final percentage. Record what evidence would move an assumption from conservative to base. A representative trial may test recovery behaviour; an order review may be more relevant to demand. [2]
Agree a stop or revise condition before placing the order. It might be a missing interface quote, insufficient demand for recovered capacity or a recovery procedure that requires skills the site cannot support. Keep the alternative of a smaller information connection or process change in the comparison.
The useful business case leaves an owner with a decision and a list of facts still worth buying. Quindart’s systems-first approach is to trace those dependencies with the people operating them, then scope the intervention and its continuing responsibilities. The downloadable worksheet keeps the calculation, evidence owner and unresolved assumptions together.
Automation ROI calculator
Separate cash savings from capacity. Test whether the expected benefit supports the investment.
Use the interactive toolYou can also export from the tool to discuss it with your team.