In-Process Controls (IPC) are the checks, tests, and monitoring activities performed during manufacturing, rather than after it, to confirm that a device or its components conform to specification at defined stages of production. In medical device manufacturing, IPC provides the objective evidence that a process stayed inside its validated operating limits.

---

What is In-Process Controls (IPC)?

In-Process Controls (IPC) sit between raw-material acceptance and finished-device release. A production route is broken into stages, and at each stage, a manufacturer defines what must be measured, how, by whom, and against which limits. Some controls measure the product: solder-joint quality on a PCBA, seal width on a sterile barrier pouch, dimensional tolerance on a molded housing. Others measure the process: reflow oven profile, melt temperature, cleanroom particle counts.

The distinction matters. Product controls tell you whether this unit conforms. Process controls tell you whether the next thousand are likely to. A well-built IPC scheme uses both, so defects are caught at the operation that created them rather than at final inspection.

---

Why In-Process Controls (IPC) matter in medical device development

For regulated devices, the stakes are not only yield. A defect that escapes to a patient-contacting or life-supporting device becomes a safety event, a field action, and potentially a recall. IPC keeps escapes rare.

There is also an evidentiary dimension. Under the FDA Quality Management System Regulation (QMSR), effective February 2, 2026, 21 CFR Part 820 incorporates ISO 13485:2016 by reference, and most old Quality System Regulation subparts are reserved. In-process acceptance is no longer cited at 820.80; it now lives in ISO 13485 Clause 8.2.6, monitoring and measurement of product, supported by Clause 7.5.1. Under EU MDR 2017/745, the same activities feed the technical documentation that a notified body reviews.

Auditors read IPC records as a proxy for process discipline. Missing signatures, out-of-tolerance readings with no disposition, or sampling plans nobody can justify statistically are common inspection findings. Weak IPC also inflates the cost of poor quality: rework found at box build costs far more than a rejected subassembly.

---

How In-Process Controls (IPC) work in a device production line

IPC is designed backward from risk:

• Identify critical parameters. Process FMEA and the risk management file (ISO 14971:2019) point to process characteristics whose failure would affect device safety or essential performance.
• Set specifications and limits. Each parameter gets an acceptance limit traceable to a design output. Limits that cannot be traced to a requirement rarely survive an audit.
• Choose the control method. Automated inline inspection (solder paste inspection, automated optical inspection, X-ray for hidden joints), operator verification, gauge measurement, or continuous process monitoring.
• Define sampling. Full inspection, attribute sampling, or variable sampling, with a documented statistical rationale. ISO 13485 Clause 8.4 requires the determination of statistical methods.
• Record the result. Readings are captured in the batch or device history record, governed by 21 CFR 820.35 and ISO 13485 Clause 4.2.5.
• Handle nonconformance. Any out-of-limit result triggers Clause 8.3, control of nonconforming product: segregation, disposition, and, where warranted, escalation to CAPA.

One relationship deserves care. ISO 13485 Clause 7.5.6 requires validation of any process whose output cannot be verified by later monitoring or measurement. Sterilization, sealing, and welding usually fall there. Where you cannot verify, you validate and then monitor process parameters. Where you can, you inspect. Most lines run both, and the boundary should be stated rather than assumed.

---

Common challenges and best practices

The most frequent failure is inspection theater: checks performed and recorded but never analyzed. Data nobody trends cannot detect drift. Feed IPC results into statistical process control charts and review them at a defined cadence.

A second problem is limits set by convention rather than evidence. If an operator must confirm a solder fillet looks acceptable, define acceptable with an image and a measurable criterion. Subjective attributes need reference standards, trained operators, and periodic attribute agreement analysis.

Third, IPC schemes drift out of alignment with change control. A tooling change, a supplier change, or a firmware update to an inspection system can invalidate the assumptions behind a sampling plan.

Good practice looks unglamorous: fewer controls, better chosen. Ten well-justified checks with tight feedback loops outperform forty nobody trends. Keep the traceability chain visible from risk file to control plan to record, because that chain is what an auditor will walk.

---

How SJML helps with In-Process Controls (IPC)

SJML builds in-process control into manufacturing programs rather than adding it afterward. Its medical PCBA lines run high-speed SMT with solder paste inspection, automated optical inspection, and X-ray inspection, while ISO Class 7 and Class 8 cleanrooms and ESD-controlled areas provide the environmental controls that surround them. Process validation (IQ, OQ, PQ), PFMEA, and PPAP establish the parameters worth monitoring, and an SAP-integrated MES captures the resulting records with unit-level traceability. QARA teams keep the control plan aligned to ISO 13485 and the QMSR as designs and suppliers change.

Talk to SJML’s manufacturing team →

---

Frequently asked questions

---

Related terms

• Process Validation
• Device History Record (DHR)
• Statistical Process Control (SPC)
• Nonconforming Product
• Batch Record