Industrial performance investment: brute capacity or flexibility?

Two camps, one budget: raw capacity vs flexibility, and the compromis that costs twice

When the order books overflow, two arguments clash: increase capacity or gain flexibility. Allocating the budget 'a bit of both' often yields mediocre results on both sides. Digital proof refocuses the debate on shared metrics and shows where every euro creates sellable throughput.

The nuance: flexibility does not necessarily oppose capacity. Reducing setups, improving sequencing or making resources versatile often increases effective capacity without adding a machine. The question becomes: which loss to eliminate before buying?

What you really pay for: fixed costs, variability, lead times, energy, and cash tied up

A new asset increases CAPEX (capital expenditure) and weighs down OPEX (operating expenses) through maintenance, consumables and parts. Flexible equipment requires additional industrialization and control. The hidden cost often comes from variability: more lot changes generate adjustments, rejects and micro-stops, hence fewer sellable parts.

Without considering WIP (work in progress) and lead time, you are financing nominal capacity instead of sellable parts. A business case must include these induced costs, not only the purchase price.

Stable bottleneck: when additional pace creates value

Factory Physics principles remind: the system does not produce faster than its constraint. A faster machine creates value if the bottleneck is identified, stable, and saturated during its uptime. Otherwise, the asset becomes inventory and disorder.

A stable bottleneck can however gain useful time through shortcut settings, targeted maintenance, or waste reduction. These levers relate to operational flexibility and increase effective capacity without increasing nominal speed.

Variability and changes of series: when flexibility wins on the 'fast' machine

If setups and mix reduce useful time, flexibility can release more sellable throughput than nominal speed. Tooling, external preparation and rapid changes reduce non-productive times. The right metric remains throughput (overall flow), not the tech sheet.

Flexibility also helps stabilize the flow: better sequencing, versatility, and reduced losses from series transitions decrease WIP and lead time. Without systemic measurement, these gains remain invisible in investment reviews.

Solving the wrong problem: WIP exploding and lead time drifting

Injecting capacity without addressing the downstream increases intermediate inventory and lengthens lead time. WIP often masks the real constraint: a fast machine running but with parts not shipped. Conversely, adding unconstrained flexibility creates complexity and lowers OEE.

The rule: a CAPEX is justified only if it increases global throughput. Measure the second-order effect before approving.

III- Trade off by replacing intuition flow simulation the protocol that

Building the baseline: cycle time, failures, yields, driving rules

Simulation starts with a baseline describing cycle time, failure rate, rejects, setup times, lot sizes and priority rules. The model must represent real variability with distributions, not smoothed averages. This calibration aligns production, maintenance, quality, and finance.

A baseline reveals invisible losses: micro-stops, queues, rework and unstable priorities. It allows testing whether fewer setups or better sequencing create more capacity than adding a machine.

Stress testing scenarios: capacity, flexibility, mix, ramp-up

The protocol simulates adding a machine, reducing setups, a change in sequencing or a ramp-up. It runs these scenarios under breakdowns, supplier delays, quality variability and absenteeism. The objective: robustness, not a best-case isolated.

Testing hybrid scenarios is essential: a small investment in flexibility can free time at the bottleneck, then give way to CAPEX if demand rises. Simulation allows phasing the effort and spotting the tipping point.

The field verdict: global throughput, overall OEE, lead times, and work-in-progress

The final decision relies on throughput (overall rate), the TRS, lead time and WIP. The simulation reveals whether reducing setup times makes a new machine unnecessary, or vice versa. The right scenario maximizes sellable parts and minimizes the need for working capital.

Summarize the decision for the committee with a before/after comparison on four indicators and a full cost, then a robustness reading. If a scenario only holds on paper, variability will cause it to collapse.

Required minimum data and a short plan to make them reliable without a “big bang”

The essential data: cycle time, failure rate, rejects, setup time, lot sizes and priority rules. The short plan combines historical extraction from the MES (Manufacturing Execution System) if present, targeted measurements on the bottleneck and field reviews to validate the hypotheses. This is enough to arbitrate without building a perfect reference.

Measuring better on few elements is better than measuring roughly everywhere. Unreliable setup times overestimate the contribution of a fast machine and undermine judgment.

Numbered steps: assumptions, gains, costs, risks, scenarios, decision

1. Demand, mix and price assumptions.

2. Baseline operational figure.

3. Expected gains expressed in EBITDA and cash.

4. Costs: CAPEX, OPEX and integration.

5. Risks.

6. Compared scenarios.

7. Financial indicators.

Decision with Go/No-Go milestones.

A solid business case for an investment in industrial performance directly links cash-flow scenarios to financial assumptions. The simulation connects shop floor to finance and shows whether a flexibility lever is better than a cadence purchase.

CAPEX vs OPEX: avoiding omissions that break ROI after signing

CAPEX covers assets, integration, works and compliance. OPEX covers maintenance, consumables, energy, licenses and additional staff. Forgetting industrialization and ramping up cadence distorts ROI during the ramp-up.

Also quantify what flexibility "consumes": methods, qualification, training and complexity of control. It can reduce hidden costs: WIP, startup rejects, overtime related to emergencies.

ROI, IRR and NPV: calculations, reading, and a simple example with round numbers

ROI (return on investment) serves as a quick comparison, the IRR (internal rate of return) compares to the cost of capital, the NPV (net present value) measures the updated value. Example: CAPEX 2,000,000

€, EBITDA gains 600,000 €/year → ROI ≈ 30% and payback ≈ 3 years without discounting.

Tier the gains in steps rather than in a full annual gain from month 1. The ramp-up often delaying benefits, this prudence avoids overestimating an investment for industrial performance.

Public aid: integrate it into net CAPEX, the schedule, and project constraints

Subsidies reduce net CAPEX but impose procedures, milestones and justifications. They affect cash flow, as the payment often arrives after completion. Always compare a "with aid" scenario and a "without aid" scenario and do not rely solely on the subsidy.

Two useful comparisons: scenario with aid vs scenario without aid

Scenario A: CAPEX 2.0 M€, EBITDA gains 0.6 M€/yr → simple payback ~3.3 years. Scenario B: 20% aid → net CAPEX 1.6 M€, payback ~2.7 years if the schedule holds. The payment timing shift reduces the impact on NPV.

V- Prioritize where to invest: KPI reading grid → families of levers → measurable effects

Industrialization and capacity: expected impact on throughput, scrap and lead time

Investing in capacity is justified when overall throughput tops out at a stable bottleneck. Industrialization (ranges, tooling, standards) reduces the gap between theoretical potential and sellable throughput.

Indicators: sellable throughput, scrap, OEE and lead time. Check absorption downstream.

Often, industrialization and flexibility converge: better-designed tooling reduces adjustments and increases effective capacity. The useful investment converts these gains into shippable parts, not flattering local indicators.

Flow and implementation: expected impact on work-in-progress, lead times and indirect productivity

An optimized layout reduces distances, handling and priority conflicts. Indicators: WIP, lead time, indirect productivity and safety. Low CAPEX on implementation often brings a rapid cash effect.

An implementation that reduces crossings lowers queues and work-in-progress. It makes production less sensitive to variances and gains effective capacity.

Automation: expected impact on variability, quality and unit costs

Automation is valuable when human repeatability limits bottleneck performance. It stabilizes quality and can reduce unit cost, but introduces dependence on maintenance and on

suppliers. The decision requires a complete scenario including capability development and maintenance plan.

Partial automation can also increase flexibility if it reduces tool-change times. Test its impact on overall throughput under variability.

Digital and steering: expected impact on decisions, stability and learning speed

Digital improves decision quality and flow stability. A digital twin allows testing governance rules without interrupting production. Indicators: schedule adherence, reduced variation, and

problem-resolution time.

Digital creates operational flexibility: better resource allocation, anticipation of shortages, and more robust sequencing. If not supported by a simulation, it risks becoming a recurring cost without impact.

Quality, maintenance and energy: expected impact on cost of non-quality, availability and kWh per unit

Reducing scrap frees capacity without new equipment. Improving maintenance increases bottleneck availability. Measuring energy in kWh per unit links performance and carbon trajectory. A decision model integrates these families to avoid conflicting trade-offs.

These levers often buy the most effective capacity at the lowest CAPEX. Their impact should appear at the same level as asset purchases in the investment review.

VI- Deploy without disruption: phased in three horizons and Go/No-Go criteria

0–3 months: diagnostic, modeling, and initial scenario tests

Short phase to build the baseline, extract histories, and define 2–3 testable scenarios. The Go rests on a model tuned to throughput, WIP, and lead time. A No-Go occurs if data remain inconsistent or if the bottleneck changes too often.

3–12 months: pilot, field validation, and measurement standard before/after

The pilot applies a scenario from the simulation with identical measurements before/after. Go criteria include increased sellable throughput and reduced WIP without degrading quality. The pilot also validates setup time, sequencing stability, and real versatility.

12–24 months: deployment, ramp-up, and financial and industrial post-audit

Deployment follows a ramp-up plan with performance milestones and integration budget. The post-audit compares real gains and assumptions on EBITDA, cash, lead times and quality, then standardizes success across other sites. Without a post-audit, the organization repeats the same mistakes.

Multi-disciplinary governance: production, finance, maintenance, quality, and purchases

Governance defines responsibilities, milestones and escalation rules. The review must decide on facts, not on narratives. The simulation model provides the common reference framework and facilitates shared decision-making.

VII- Mini-cases: real arbitrations, quantified impacts, success conditions

The figures below correspond to orders of magnitude frequently observed by Dillygence during diagnostic, simulation or industrial transformation missions, in comparable contexts. They do not constitute a promise of results and serve to illustrate mechanisms of

decision. The real impacts depend on the product mix, variability, and the level of on-site control.

VIII- The traps that ruin profitability, and the field counter-measures

Confuse machine throughput and sellable capacity

Trap: measure throughput rather than conforming shipped pieces. Countermeasure: link each CAPEX to an identified bottleneck and manage the sellable throughput and the cost of

non-quality. Verify by simulation before purchase.

Underestimate ramp-up, interfaces, and integration costs

Trap: machine budget without integration budget. Countermeasure: ramp-up plan, integration budget and Go/No-Go milestones. Test degraded scenarios in the simulation.

Optimize locally and move the bottleneck

Trap: local gains generating new downstream constraints. Countermeasure: systemic reading followed by validation by global throughput and work-in-process. Integrate upstream and downstream into the model.

Neglect field organization: skills, maintenance, and standards

Trap: asset delivered without skills. Countermeasure: training plan, available parts and clear responsibilities. Success is played out on the field.

Not measuring afterwards: no post-audit, no learning

Trap: no post-deployment review. Countermeasure: financial and industrial post-audit with indicators identical to the baseline. Turn gaps into corrective actions.

Conclusion

An investment for industrial performance is not judged by the nominal speed of a machine, but by the increase in conforming parts shipped and the decrease in WIP (work in progress, in-process, production) and lead time. Buying capacity without addressing the flow often leads to more stock, more conflicting priorities and a diluted ROI during ramp-up. Quantify, via a digital twin, what every euro of CAPEX truly transforms into saleable throughput and cash: this is the only way to make an investment in industrial performance defendable in committee and sustainable on the field.

Dillygence combines field expertise and a digital twin to compare your CAPEX and flow scenarios on measurable results, then turn proof into durable industrial and financial gains.

FAQ — Investments and Industrial Performance

What is industrial performance?

Industrial performance is the ability to deliver conforming parts, at the right cost and on time, with controlled energy consumption and CO2 emissions. It combines saleable throughput, quality, lead time, unit cost, safety and stability. Performance is measured at the level of the full flow, not a single machine.

What is the investment to improve industrial performance?

It is an expense that increases saleable throughput or durably reduces system losses. It may target capacity, flexibility, flows, quality, maintenance, energy or digital control. Its validity comes from a business case linked to field KPIs and validated by simulation or digital twin.

Which indicators to track to manage an investment?

Track sellable throughput, TRS, lead time, WIP, service rate, scrap, kWh per unit, CAPEX, OPEX, and the impact on working capital requirements. Manage the gaps between assumptions and actual results during ramp-up.

Which levers to fund first?

Prioritize the lever that relieves the constraint identified by simulation. Stable bottleneck → capacity. High variability → flexibility or reduction of setups. High non-quality → quality. Flow instability → layout and control rules.

How to reduce the financial risks of an investment?

Limit risk through simulation, targeted pilots, and Go/No-Go milestones. Test the systemic effect on overall throughput, WIP, and lead time. Require post-audit and contractual clauses on performance ramp-up. The main risk remains shifting the constraint or overestimating gains during ramp-up.

How to standardize the approach across multiple sites?

Define a baseline of minimum data, a simulation protocol, a business case model, and a measurement standard before/after. Adapt local parameters but keep the same indicator definitions.

How to assess the long-term impact on capacity and lead time?

Project overall throughput and work-in-progress under demand, breakdown, and product-mix scenarios. Verify that the increase in capacity does not generate an increase in WIP and lead time. Confirm with a post-audit after deployment.