ROI calculator

Calculate the ROI of your industrial project

Estimate the value of an AI project across your real levers. Productivity, maintenance, quality, production, energy, supply chain, engineering. A conservative calculation, by use case, with no double counting.

Calculate my ROI

Free · No sign-up · Instant result

Value trajectory

Cumulative value overtakes the investment, then the gap widens.

InvestmentBreak-evenCumulative valueGo-live3 years
Cumulative valueInvestment

The essentials

  • The formula stays simple: ROI = (gains - costs) / costs × 100.
  • The calculation distinguishes three natures of value. Direct savings (cash), revenue or margin (cash if sellable), recovered capacity (non-cash, shown separately).
  • The benchmarks stay below the maximums published by Deloitte, McKinsey and the World Economic Forum. Reference points, to confirm with your figures.
  • The engine removes double counting: two gains that overlap are not added in full.

Calculate your AI project ROI

A four-step path: your industry, your levers, your figures, then your costs. The result updates live. The calculation stays on your screen.

Your industry

It tailors the use cases shown in the next step. It does not change the figures in the calculation.

Your value levers

Choose 1 to 3 levers. The calculation only combines what you tick.

0 / 3 selected

Your operating data

We prefill conservative benchmarks. Replace them with your figures to make the estimate more reliable.

Your project costs

An honest ROI includes the full project cost, beyond the licence alone.

Year 1 (non-recurring)

Recurring (each year)

The first year includes the first annuity of the recurring costs.

See my result
Discuss it with Assets 4.0

We refine these figures with you, free and with no obligation.

What ROI can AI generate in industry?

Gains depend on the lever, the sector and maturity. Here are external benchmarks, published and verifiable.

LeverObserved (sources)Benchmark used hereSource
Productivity and automation+7 to +20% productivity+5 / +10 / +15%Deloitte, Smart Manufacturing 2025
Productionup to +20% volume+3 / +7 / +10%Deloitte, Smart Manufacturing 2025
Freed capacityup to +15% capacity+3 / +7 / +10%Deloitte, Smart Manufacturing 2025
Maintenance and reliability-30 to -50% downtime-10 / -20 / -30%McKinsey, predictive maintenance
Quality-41% defects (lighthouse sites)-10 / -20 / -30%World Economic Forum
Energy-10 to -25% consumption-3 / -7 / -10%Industrial cases, ScienceDirect
Supply chain (inventory)-20 to -30% inventory-5 / -10 / -15%McKinsey, forecasting
Engineering / R&D+10 to +15% productivity+5 / +10 / +15%McKinsey

The simulator applies conservative values, below these benchmarks. It keeps a reasonable central benchmark, adjustable at any time.

The 8 value creation levers

Each lever turns an operating figure into annual value. You combine one to three of them. The engine does not add two gains that overlap.

Productivity and automation

Recoverable repetitive hours, valued at the loaded hourly cost.

AI Systems Design

Maintenance and reliability

Downtime hours avoided, valued at the hourly downtime cost.

Maintenance Intelligence

Quality

Reduction of the cost of poor quality: scrap and rework.

Food safety & compliance

Production

Extra capacity that is truly sellable, valued at contribution margin.

AI Systems Design

Energy

Avoidable share of consumption on the annual energy spend.

AI Systems Design

Supply chain and inventory

Inventory freed and carrying cost avoided on the stock.

AI Systems Design

Engineering

Productivity recovered on study and design hours.

AI Systems Design

Compliance and audits

Documentation time and external services avoided.

AI opportunity audit

How do you calculate the ROI of an AI project?

ROI compares the value created to the cost incurred. The formula stays simple.

Formula and indicators

  • Year 1 ROI = (gains - costs) / costs × 100.
  • Example: €120,000.00 of gains for €60,000.00 of costs gives a 100% ROI.
  • Payback: the initial investment divided by the net monthly gain. It is expressed in months.
  • Three-year net value: the gains added up, with a 60% ramp-up in the first year. The total cost over three years is then subtracted. Amounts are not discounted.

Before any calculation, the engine distinguishes four natures of value.

  • Direct savings: cash actually avoided.
  • Recovered capacity: the value of the time made available, shown separately.
  • Additional revenue: counted when the capacity is sellable.
  • Risks avoided: shown separately, because they are probabilistic.

Which costs should you include? (TCO)

An honest ROI includes the total cost of ownership. The simulator asks for seven items.

  • Audit and scoping (year 1).
  • Development and integration (year 1).
  • Team training (year 1).
  • Internal project time (year 1).
  • AI and software licences (recurring).
  • Infrastructure (recurring).
  • Maintenance and support (recurring).

The result shows the total cost of ownership in year 1, first annuity of the recurring costs included. It then shows the cumulative cost over three years.

Examples of AI ROI in industry

Three illustrative estimates, with conservative values consistent with the simulator. Your figures remain the only reference.

Integrity Loop dashboard
ROI shows up in indicators, not in promises. Here the asset dashboard in Integrity Loop, fed by the reports you already receive.

Predictive maintenance, food & beverage

Unplanned downtime
200 h/yr
Downtime cost
€3,000.00/h
Reduction used
-20%
Annual value
€120,000.00
Year 1 investment
€60,000.00
~100%Payback ~5 months

Quality, pharmaceuticals

Cost of poor quality
€800,000.00/yr
Reduction used
-20%
Annual value
€160,000.00
Year 1 investment
€120,000.00
3-year net value
~€236,000.00
~33%Payback ~8 months

Productivity, machining

People involved
25
Repetitive hours
10 h/wk
Recovered share
10%
Annual value
~€58,500.00
Year 1 investment
€38,000.00
~54%Payback ~7 months

These examples rely on conservative reductions (10 to 20%), below the published maximums. They illustrate a method, to confirm with your data.

Methodology and sources

Our benchmarks come from public, verifiable sources. We deliberately stay below the observed maximums.

Deloitte, Smart Manufacturing survey 2025

Up to +20% production, +20% workforce productivity and +15% freed capacity.

deloitte.com
McKinsey, maintenance and reliability

Predictive maintenance reduces machine downtime by 30 to 50%. It extends equipment life by 20 to 40%.

mckinsey.com
McKinsey, supply chain forecasting

AI applied to forecasting reduces errors by 20 to 50%. Inventory falls by 20 to 30%.

mckinsey.com
World Economic Forum, Global Lighthouse Network

AI use cases at lighthouse sites reduce defects by about 41%. Energy consumption falls by about 28%.

weforum.org
ScienceDirect study, AI and factory energy

AI-based energy optimisation reaches 10 to 25% consumption savings in industry.

sciencedirect.com

A reliability indicator comes with each estimate. It states the share drawn from your data.
Updated on 5 August 2026. Method and calibration: the Assets 4.0 team.

Frequently asked questions

How do you calculate the ROI of AI?
ROI is calculated as follows: (gains - costs) / costs × 100. Gains combine direct savings and recovered capacity. Costs cover audit, integration, licences, infrastructure, training and support. Add the payback period and the net value over three years.
What is the average ROI of AI in industry?
There is no universal average ROI. Benchmarks vary by lever. Productivity rises by 7 to 20% (Deloitte). Downtime falls by 30 to 50% (McKinsey). These figures are observed maximums, to confirm on your site.
How do you calculate the ROI of predictive maintenance?
Multiply your unplanned downtime hours by their hourly cost. Apply a conservative reduction, for example 20%. Compare this gain to the full project cost. McKinsey observes 30 to 50% fewer stoppages.
Which costs should you include in an AI project?
Include seven items: audit and scoping, development, licences, infrastructure, training, internal time and support. Separate year 1 from recurring costs. You get a realistic total cost of ownership over three years.
How long does it take to pay back an AI project?
Payback depends on the lever and the cost. A focused project often pays back in six to eighteen months. Compute it by dividing the investment by the estimated monthly gain.
Which AI use cases are the most profitable in industry?
Predictive maintenance and quality often deliver the fastest returns. Productivity and energy follow. Extra production only counts if it is truly sellable.

From estimate to business case

We refine these benchmarks with your real data. Then we scope an AI project with a quantified ROI. Free and with no obligation.

Book a call