EdgeBits ROI Calculator

Your plant

All values annual unless noted. Numbers in ₹.

Number of plants (sites you'd deploy to)

Annual revenue per plant (₹ Crores)

Downtime cost per hour (₹ Lakhs / hr, all-in: lost margin + fixed costs still running + labour)

Rule of thumb: 2 to 4 times revenue-per-production-hour. For a ₹200 Cr plant on 2 shifts, revenue-per-hour is about ₹50,000, so a realistic downtime cost sits in the ₹1 to ₹2 L range. Higher for continuous processes with customer penalties.

Unplanned downtime (hours per month, across plant)

Typical mid-market plants log 40 to 80 hours/month of unplanned downtime. World-class TPM programs sustain 4 to 8 hours/month.

Annual energy bill per plant (₹ Crores)

Paying a meter / historian / MES vendor to hold your data? Vendor-hosted dashboards, per-meter or per-tag hosting fees, per-user portal seats. Data goes to their server; EdgeBits keeps it on yours.

Currently paying for a historian? AVEVA PI, FactoryTalk Historian, Wonderware, Ignition — Edge Analytics replaces the licence line.

Manual shift reports, ERP feeds, or energy logs today? Shift incharges typing daily production numbers into Excel or SAP.

Continuous process (pharma, chemical, FMCG, food)? Adds a scrap / rework reduction estimate for process-drift catches.

Your OEE

Speak the language your TPM manager already uses. Availability × Performance × Quality.

Number of plants

Annual revenue per plant (₹ Crores)

Current Availability (A — run time ÷ planned time)

30%78%100%

Current Performance (P — actual cycle ÷ ideal cycle)

30%82%100%

Current Quality (Q — good count ÷ total count)

70%97%100%

78%

82%

97%

OEE

62%

Your Top Loss category (where 50%+ of your loss hours sit)

Manual reason-code capture today? Operators write reasons on a clipboard next shift. Adds a 100%-capture uplift + Top-Loss-report FTE savings.

Currently paying for a historian? AVEVA PI, FactoryTalk, Wonderware — Edge Analytics replaces the licence.

Your annual savings

Live: updates as you type on the left.

Year-1 savings estimate

₹0

Across the five buckets below.

Year-1 numbers. Steady state (18–24 months) is usually 1.5–2× this. Plants that install and don't act on alerts save close to 0%. Range depends on baseline maturity, sector, and demand mix — see "How we compute this" below.

Breakdown by bucket

Downtime reduction ₹0

Energy savings ₹0

Vendor-hosted dashboard eliminated ₹0

Historian licence eliminated ₹0

Manual reporting saved ₹0

Quality & scrap reduction ₹0

Typical payback: 6 to 12 months at this savings profile. Exact payback depends on plant size, deployment scope, and which buckets you prioritise. We'll walk through it in the call.

Get a custom proposal See how "Ask Your Plant" works

How we compute this

This calculator is a first-order estimate, not a proposal. The multipliers below sit in the conservative half of the range reported by 2025 industry benchmarks. Every number cited here has a URL you can check. We treat "median-of-vendor-claims" as a red flag, not a source.

What the numbers below represent: year-1 outcomes for a typical mid-market industrial plant that (a) implements EdgeBits, (b) uses the data, and (c) has a maintenance / operations team that acts on the alerts. Steady-state results at 18–24 months are usually 1.5–2× these numbers. Plants that install and don't act on alerts save close to 0%.

General mode · what we use

Downtime reduction 25–35% of unplanned downtime avoided, year 1. Industry range: 30–50% (2026 predictive-maintenance benchmark), 35–45% (US DoE data), up to 70–90% at maturity. We use 25–35% because year 1 rarely hits the ceilings.

Energy savings 5–10% of annual energy bill, year 1. Industry range: 8–25% (2025 monitoring guide); 15–25% with structured programs (sub-metering ROI). We use the low half because sub-metering alone lands ~8%; structured programs take a year to reach 15%+.

Vendor-hosted dashboard eliminated ₹5 L per plant per year, only if you currently pay a meter / historian / MES vendor for a hosted dashboard or portal. Meter-supplier SaaS: typically ~₹1000–1500 per meter per month + per-user portal seats. Twelve meters + a handful of users lands ~₹3 L/yr; twenty-plus meters or multiple sites scales linearly. See Take back your energy data for the worked math. If your dashboard is Grafana / Power BI on infrastructure you already own, untick this.

Historian eliminated ₹15 L per plant per year, only if you currently pay for AVEVA PI, FactoryTalk, Wonderware, or Ignition. Real range varies enormously by tag count and vendor. If you use Grafana + InfluxDB (free), untick this; the saving doesn't apply.

Manual reporting saved 0.4 FTE = ₹8 L/yr, only if shift reports, ERP feeds, or energy logs are typed by hand today. Automated plants: untick.

Quality & scrap 0.3% of annual revenue, only for continuous processes (pharma, chemical, FMCG, food). Genuinely conservative floor. Industry range is 1–3% at steady state; we don't claim that year 1.

TPM · OEE mode · what we use

Availability lift +3 to +5 pp year 1. Industry range: 3–8 pp typical (see 2026 OEE benchmarks). Steady state after 18–24 months is often 2× this.

Performance lift +2 to +4 pp year 1. Once Small Stops and Slow Running are visible in the data.

Quality lift +1 to +2 pp year 1 on plants already > 95% Q. SPC bands on CTQ variables. Bigger lifts happen only when the current Q floor is below 90%.

A / P / Q are not fully independent Fixing Small Stops moves hours between the A and P buckets rather than adding to both. The calculator applies a 15% coupling discount to reflect this; naive addition would over-estimate.

Top-Loss boost 1.3× on the component that owns your declared Top Loss. Focused Improvement teams reliably outperform generic-lift assumptions on the one loss they attack. No academic single-number source; this is directional.

OEE → ₹ conversion New output value = revenue × (new OEE ÷ old OEE − 1) × 12% contribution margin. Standard TPM finance framing. See caveat below.

Reason-code capture ₹6 L/yr per plant. Manual logs miss ~30% of stop events; automated capture ≥ 50% more accurate (2025 downtime-tracking benchmark). Poor OEE data quality costs plants 8–15 pp of recoverable efficiency (iFactory 2025).

Historian eliminated ₹15 L/plant/year. Same as general mode.

Two honest caveats

Contribution margin varies by sector. Default 12% (mid-market industrial). FMCG and pharma see 20–40%; steel and commodity plants see 5–15%. Your proposal uses your actual margin.

The OEE-to-₹ math assumes you can sell the extra output. Demand-limited plants (order book full at current pace) convert OEE lift to cost reduction rather than revenue growth. Same rupee number, different lever.

Not modelled: EdgeBits subscription + implementation cost, ramp-up curve, and implementation risk (~30% of industrial-IoT pilots stall). This is a first-order savings estimate, not a payback calculator.

Full research trail with 2025–2026 source URLs is in issues/in-progress/website-roi-calculator-research.md. Every multiplier ties back to a citation there. If a number changes here, that file changes first.

Abbreviations used on this page

pp — percentage points. The absolute gap between two percentages. Availability going from 78% to 82% is a +4 pp lift, not a "+4%" lift.

FTE — Full-Time Equivalent. One person working full-time for a year. "0.4 FTE" = the labour hours of one person 40% of the time (roughly 2 days a week).

OEE — Overall Equipment Effectiveness. A × P × Q. The one number TPM programs live by.

A · P · Q — Availability × Performance × Quality. The three components of OEE.

TPM — Total Productive Maintenance. Nakajima's 1988 framework; still the industry standard.

CTQ — Critical To Quality. A process variable whose drift moves the defect rate (temperature, fill weight, coating thickness, viscosity, conductivity).

SPC — Statistical Process Control. Continuous monitoring of CTQ variables against control bands.

₹ L / ₹ Cr — Lakh (1 L = 100,000) / Crore (1 Cr = 10,000,000). Indian numbering.

See what EdgeBits could save your plant