Calibration vs Verification vs Validation — What's the Difference? (ISO 9001 Explained)

Quick Definitions — Calibration, Verification, Validation

Calibration — What It Is and When You Need It

Calibration is the activity of comparing a measuring instrument's output against a measurement standard whose accuracy is known and traceable to SI units through a national metrology institute (NPL India in this case). The purpose of calibration is to determine the instrument's accuracy — how close its readings are to the true value — and to quantify any error.

Calibration does not fix or adjust an instrument — it measures how accurate it currently is. After calibration, the instrument may be adjusted (if adjustment is needed and possible), but the calibration certificate documents both the as-found (pre-adjustment) and as-left (post-adjustment) states.

• Example 1 — Manufacturing: A vernier caliper used for final inspection is calibrated against a NABL-traceable gauge block. The calibration certificate shows the caliper reads 0.03 mm high at 50 mm — meaning all parts measured at 50 mm were actually 0.03 mm larger than the reading indicated.
• Example 2 — Pharma: An analytical balance is calibrated using certified reference weights. The certificate shows the balance reads 0.2 mg low at 100 g — meaning all product batches weighed at 100 g were actually 100.0002 g.
• Example 3 — Process: A pressure transmitter is calibrated against a NABL reference pressure gauge. The transmitter reads 2 kPa high at 500 kPa — all process pressure readings at 500 kPa were actually 498 kPa.

Verification — What It Is and When You Use It

Verification is the activity of confirming that a specified requirement has been met — checking whether something conforms to its specification. Verification does not require traceable measurement standards in the same way calibration does. The instrument used for verification must be calibrated, but verification itself is about the outcome of a measurement, not the instrument's accuracy.

• Example 1 — Incoming material: A batch of raw material is received and its diameter is measured with a calibrated vernier caliper to check it is within the 25 ± 0.1 mm specification. The measurement confirms conformance. This is verification.
• Example 2 — In-process: During machining, a dimension is checked every 10 parts to verify it remains within tolerance. The act of checking is verification.
• Example 3 — Software: A calculation in an Excel spreadsheet is verified by running it against known inputs and confirming the output is correct. This is design verification (ISO 9001 Clause 8.3.4).
• Example 4 — Supplier: A supplier's drawing is verified against customer requirements before approval. This is verification of design input.

Validation — What It Is and When You Use It

Validation is proving that a process, method, system, or product consistently produces the intended result under real-world conditions. Validation is required when the output of a process cannot be verified by subsequent monitoring or measurement — meaning defects are not detectable until the product is in use (or until the patient has been harmed, in medical/pharma contexts).

• Example 1 — Heat treatment: A furnace heat treatment process is validated by proving it consistently achieves the required hardness in the product. You cannot inspect hardness of every part without destroying it — so the process must be validated to give confidence in the outcome.
• Example 2 — Welding: A welding procedure is validated (Welding Procedure Specification + Procedure Qualification Record) before use in production — because you cannot detect internal weld defects in every finished weld without destructive testing.
• Example 3 — Pharma sterilisation: An autoclave sterilisation cycle is validated by proving it consistently achieves F0 ≥ 8 (12-log reduction of Bacillus stearothermophilus spores) across the entire load. You cannot test every vial for sterility without consuming it.
• Example 4 — Software: A laboratory information management system (LIMS) is validated by running it through IQ/OQ/PQ testing to prove it consistently processes and stores data correctly.

Common Confusions — and How to Clear Them

Frequently Asked Questions