A PAT-aligned framework for installing and operating particle counting systems to detect pre-limit particle-size...

ISO 14644-1 defines ISO-8 cleanrooms, often found in pharmaceutical manufacturing facilities, by maximum allowable concentrations of airborne particles at specified threshold sizes (commonly ≥0.5 μm and ≥5.0 μm) within a classification particle size range of 0.1-5 μm. For ISO Class 8, though ISO 14644- 1 lists sizes from 0.1 to 5.0 μm, classification commonly uses the operationally relevant limits of ≥0.5 μm and ≥5.0 μm. While ISO classification establishes a compliance boundary, it does not fully leverage modern optical particle counter (OPC) time series data to detect early environmental degradation driven by controllable operational factors in non-sterile areas (traffic, door cycling, material handling, and housekeeping drift). FDA's Process Analytical Technology (PAT) guidance combined with Quality by Design (QbD) principles, promotes timely measurement, enhanced process understanding, and multivariate/statistical approaches to manage variability proactively. This paper proposes a PAT aligned framework for installing and operating particle counting systems in ISO-8 non-sterile controlled areas to detect pre-limit particle size distribution (PSD) shifts. The approach integrates (i) risk-based monitoring network design, (ii) PSD-shift features (coarse to fine ratios, event rate, recovery time, persistence) that translate raw counts into actionable change signals, and (iii) statistical change detection (EWMA/CUSUM) to distinguish transient events from sustained shifts. One thing which is often overlooked is physical collection issues, especially when pulling samples through long tubes or curved lines. Particles at 5 microns or more tend to drop out under those conditions, skewing results noticeably. So this guidance covers considerations for choosing particle counters and where to place ports for truer readings. Instead of waiting until thresholds breach, the system proposed in this article acts earlier using feedback loops tied directly to control actions there by providing a robust continuous monitoring system.