How to Calculate OEE in Indian Factories: Formula, Benchmarks and a Worked Example (2026)

OEE (Overall Equipment Effectiveness) is calculated as Availability × Performance × Quality. Availability is Run Time ÷ Planned Production Time, Performance is (Ideal Cycle Time × Total Count) ÷ Run Time, and Quality is Good Count ÷ Total Count. A machine running 390 minutes out of 450 planned, producing 310 parts against a 1-minute ideal cycle with 300 good, scores 86.67% × 79.49% × 96.77% = 66.7% OEE. The globally recognised world-class benchmark is 85%, while most Indian SME factories currently sit between 45% and 62%.

What OEE actually measures

OEE answers one question: of the time you planned to produce good parts at full speed, what percentage did you actually achieve? It compresses three separate failures — stopping, slowing and scrapping — into a single number defined under ISO 22400-2, the international standard for manufacturing KPIs.

The reason it matters for Indian manufacturers is comparability. A plant manager saying “production was good this month” is an opinion. An OEE of 58% rising to 64% is a fact you can put in front of a customer, a lender or a board.

Notice that 85% is not 85% of a perfect day — it is 90% × 95% × 99%. Because the three factors multiply, a small slip in each compounds fast. That multiplication is exactly why factories that feel busy still score in the fifties.

A worked OEE calculation, step by step

Take a single CNC machine on an 8-hour shift in a Chennai auto-components unit. Here is every number you need and where it comes from.

Availability = 390 ÷ 450 = 86.67%

Performance = (1.0 × 310) ÷ 390 = 79.49%

Quality = 300 ÷ 310 = 96.77%

OEE = 0.8667 × 0.7949 × 0.9677 = 66.7%

Read the three factors, not just the total. Here, Performance at 79.49% is the weakest link — the machine was running but slower than its rated speed. Chasing breakdowns would be the wrong fix; the real money is in minor stops and speed loss.

OEE benchmarks: where do Indian factories actually stand?

The 85% world-class figure comes from Seiichi Nakajima’s TPM work and remains the reference target for discrete manufacturing. Real-world medians sit lower. TeepTrak’s 2026 benchmark data puts discrete manufacturing at 65–75% and automotive Tier-1 at 75–85%.

The spread is the story. India’s Tier-1 plants match global benchmarks, while much of the Tier 2–3 supply chain starts near 47% — largely because most of India’s MSME manufacturers still run without real-time production visibility. The gap is not a skill gap. It is a measurement gap.

The Six Big Losses behind every low OEE score

Nakajima’s TPM framework categorises everything dragging OEE below 100% into six losses. Mapping your losses to these categories tells you which factor to attack.

• Equipment breakdowns (Availability) — unplanned failures. The target of predictive maintenance.
• Setup and changeover (Availability) — attacked with SMED and better scheduling.
• Idling and minor stops (Performance) — jams, sensor trips, short material waits. The most under-recorded loss in Indian plants.
• Reduced speed (Performance) — running below rated cycle because of wear, tooling or operator caution.
• Startup and yield losses (Quality) — scrap produced while the process stabilises after a changeover.
• Production defects (Quality) — rejects and rework during steady-state running.

Minor stops deserve special attention. When downtime is logged manually, operators record the 40-minute breakdown but never the 4-minute jam that happens eleven times a shift — even though the second one costs more. This is the single biggest reason manual OEE figures come out flattering and wrong.

Manual OEE vs automated OEE: why the numbers disagree

Most Indian SMEs start with a paper logbook or an Excel sheet filled in at the end of the shift. It is a reasonable starting point and it is always optimistic. Reconstructed-from-memory data misses micro-stoppages, rounds times to the nearest ten minutes, and quietly reclassifies unplanned downtime as planned.

Expect your OEE to drop when you automate measurement. That fall is not a regression — it is the first time you are seeing the truth, and it is the baseline every genuine improvement gets measured against.

How to automate OEE measurement on your shop floor

Automated OEE needs three signals per machine: is it running, how many parts has it made, and how many were good. Everything else is derived arithmetic.

On modern equipment those signals already exist inside the PLC or CNC controller and can be read over standard industrial protocols — the choice between them is covered in our guide to OPC UA vs MQTT vs Modbus. On older machines with no controller access, a simple retrofit works: a current sensor on the motor to detect running state, plus a proximity or optical sensor to count parts. Most Indian factories run a mix of both, which is why Odoo IoT integration matters more than machine age.

Where the calculation runs is the next decision. Counting and stop-detection belong at the edge for reliability during network drops, while shift and monthly trends belong in the cloud — the trade-offs are laid out in edge vs cloud computing for Industrial IoT.

Why OEE belongs in your ERP, not a standalone dashboard

An OEE screen on the wall creates awareness. An OEE figure inside your ERP creates decisions. When effectiveness data sits alongside work orders, material availability and dispatch commitments, you can see that the 12% availability loss last week traced to a raw-material wait, not the machine — and act on procurement instead of maintenance.

That link is built with ERP API integration and surfaced to supervisors through a custom Odoo mobile app. If you are still deciding which system to implement first, our comparison of MES vs ERP vs SCADA vs IIoT sets out the sequence.

OEE vs TEEP: the number that shows your hidden capacity

OEE measures only the time you planned to run. TEEP (Total Effective Equipment Performance) measures against all 24 hours of the calendar, exposing capacity you own but do not use. In our worked example, 300 good parts against a theoretical 1,440 available minutes gives a TEEP of 20.8%.

That gap matters commercially. Before signing off on new machinery, a factory running 66.7% OEE on a single shift should ask whether an OEE improvement plus a second shift delivers the same output for a fraction of the capital. TEEP turns that from a hunch into a calculation.

How to improve OEE: what to fix, in what order

Once you know your three factors, improvement stops being guesswork. Fix the weakest factor first — the multiplication means a 10-point gain on your worst number beats a 3-point gain on your best.

If Availability is lowest

Split your downtime into breakdowns versus changeovers, because the fixes are completely different. Breakdowns are a maintenance-strategy problem solved by condition monitoring and planned intervention. Changeovers are a method problem solved by SMED — separating internal setup work that requires the machine stopped from external work that can be prepared while it still runs. Many Indian job-shops running high-mix, low-volume orders lose more time to changeovers than to failures, and never notice because both are logged simply as “machine down”.

If Performance is lowest

This is almost always minor stops and speed loss, and it is where automated measurement pays for itself immediately. Once every 3-minute jam is timestamped, patterns appear fast: a particular material grade, a specific tool nearing end of life, or one shift consistently running conservative feed rates. Fixing the top three recurring micro-stoppages typically moves Performance more than any capital purchase.

If Quality is lowest

Separate startup scrap from steady-state defects. Heavy startup losses point at changeover procedure and first-article checks; steady-state rejects point at tooling wear, process drift or incoming material variation. Quality losses are the most expensive of the three because you paid for the material, the machine time and the labour, then threw the result away.

What one OEE point is worth in rupees

OEE only earns board attention when it is translated into money. The conversion is straightforward: one OEE point equals 1% of your planned production time converted into saleable output, so its value equals 1% of the contribution margin that line could generate at full planned utilisation.

Take the worked example. At 66.7% OEE the machine produced 300 good parts per shift. Moving to 72% — a realistic first-year gain from removing minor stops — yields about 324 parts from the same shift, the same labour and the same electricity. Those 24 extra parts a shift carry almost no marginal cost, so nearly the entire contribution margin drops to the bottom line. Across 300 working days, that is 7,200 additional parts from capacity you already own and already pay for.

Present it that way and the conversation changes. OEE stops being a production metric that engineers care about and becomes a capital-avoidance argument the owner cares about.

Four mistakes that make OEE useless

1. Inflating the ideal cycle time. If you set the ideal rate to what the machine usually does instead of its nameplate best, Performance flatters you permanently. Use the fastest sustained rate the machine has genuinely achieved.
2. Reclassifying downtime as planned. Moving a chronic material shortage into “planned stop” raises OEE without producing a single extra part. Be strict about what counts as planned.
3. Counting rework as good. Quality means good first time. Parts that needed rework consumed capacity twice.
4. Using OEE to judge operators. The fastest way to kill data honesty is to tie it to individual appraisals. OEE measures the system, not the person standing at the machine.

A 30-day plan to start measuring OEE properly

1. Week 1 — define. Pick one bottleneck machine. Agree the ideal cycle time, what counts as planned downtime, and what counts as a good part. Write the definitions down.
2. Week 2 — measure manually. Log stops, counts and rejects by hand for one week. Crude, but it produces a first baseline and reveals which losses dominate.
3. Week 3 — instrument. Fit run-state and part-count sensors, route them through a gateway, and compare automated numbers against the manual week. The gap is your measurement error.
4. Week 4 — act on one loss. Attack the single largest loss category. Re-measure. A 5-point OEE gain on a bottleneck machine is a credible first result.

Scale to the next machine only after the first one is trusted. OEE fails in Indian factories far more often from disputed definitions than from missing technology.