For years, in many companies, cleaning quality has been assessed using a simple yet risky criterion: "clean to sight." It works... until it doesn't. All it takes is an audit, a supplier change, a more stringent customer requirement, or a downstream issue (painting, assembly, sealing, gluing) to discover that "it looks clean" doesn't equate to "it's clean to specification.".
The good news is that there are practical approaches to moving towards objective and repeatable criteria, without turning the department into a laboratory. The goal isn't to measure everything, all the time: it's to measure what matters, with consistent methods and indicators useful for production and quality.
The degree of cleanliness is not a universal value. It varies based on three variables.:
A oleodynamic or automotive component may require inspections for particulate contamination (microparticles that compromise valves, pumps, and circuits). In these cases, specific technical standards exist, such as ISO 16232 and VDA 19.1, designed specifically to inspect and document the technical cleanliness of functionally critical components.
In controlled environments (e.g. cleanrooms) cleaning can also be linked to particle classifications and monitoring of the environment.
Oil, grease, shavings (chips), dust, abrasive pastes, processing residues, soluble salts: each contaminant requires a different control logic (visual, gravimetric, particulate, chemical).
The same part may be "clean enough" for assembly, but not for painting or gluing. The quality of cleaning must be determined based on the downstream risk.
The starting point, therefore, is a simple question: "What happens if this contaminant remains on the part?" From there, a measurable specification is defined.
Here it is best to think in terms of levels: rapid checks (frequent) + objective checks (sample), so as to combine operability and robustness.
These are quick checks, to be done in the department, useful for intercepting deviations.
Startup and shift change checklist: filter status, bath level and conditions, free nozzles, cycle parameters (time/temperature/pressure), correct load;
Guided visual inspection: non a occhio, ma con criteri chiari (punti critici del pezzo, illuminazione, distanza, limite accettabile);
Deviation indicators: simple signals that say “we are going out of range” (e.g. increased times, increased rework, recurring residues in an area).
These checks do not certify cleanliness, but they prevent the process from going astray without anyone noticing.
When it is really necessary to measure, more robust methods come into play, to be applied with a defined frequency (daily/weekly/monthly depending on the risk).
Particle contamination (technical cleanliness): Particle extraction, filtration, and analysis (counting/sizing). This is the typical approach for critical components in the automotive and related industries, with references such as ISO 16232 and VDA 19.1.;
Non-volatile residue (NVR) / gravimetry: useful when the interest is to quantify "how much residue remains", especially for oily films or non-visible contaminations;
Test for soluble salts (when relevant): Before coatings or surface treatments, the presence of soluble contaminants can be critical; there are standardized methods such as the ISO 8502 family (e.g., the Bresle method for extraction and analysis);
ATP bioluminescence (in specific contexts):Widely used in the food industry for rapid hygiene checks; useful as an indicator of organic residues, with limitations and interpretations that must be managed correctly.
Rule of thumb: Choose a couple of methods consistent with your industry and quality control, and build simple controls around them. Better a little but repeatable, than a lot and inconsistent.
The goal of KPIs is to anticipate problems before they become a downtime or a complaint.
If you don't want to complicate operations, avoid time-consuming KPIs. Focus on indicators that speak the language of production and quality.:
Percentage of batches compliant with the first pass (first-pass wash yield);
Deviations of process parameters (time/temperature/consumption) from the target range;
Trends in filter clogging/nozzle cleaning/bath replacements.
Parts cleaned/reworked (hours and cost);
Rejections attributable to contamination;
Complaints or nonconformities reported by the customer related to residues.
Average cycle time (and variability);
Machine availability (maintenance/clogging downtime);
Specific consumption (water/energy/detergent per batch or per kg of parts).
An effective control plan has four characteristics: clear, light, repeatable, improvable.
Identify the most critical parts or families (by function or downstream process);
Define the enemy contaminant (particles? Oily film? Salts? Organic residues?);
Translate everything into a specification: limit, method, frequency, responsibility.
Every shift: Quick checklist + guided visual inspection;
Weekly (or depending on risk): Objective random testing (1 primary method);
Monthly/quarterly: KPI review + internal process audit.
Many plans fail not because of the chosen method, but because the conditions change:
Inconsistent sampling (different areas of the part);
Varying times between washing and testing;
Post-wash handling/contamination;
Misinterpreted acceptance criteria.
This is where simple procedures and training make the difference.
When an indicator goes out of range, the response must be immediate and feasible, with activities such as:
Cleaning nozzles/replacing filters;
Adjusting cycle parameters;
Restoring detergent concentration/bath conditions;
Reviewing the load and racks/positioning parts.
Remember that the goal is to monitor the process and make decisions that can concretely improve the performance of your wash cycle.
Measuring cleaning quality is only possible if the process is repeatable. And repeatability doesn't happen by chance: it depends on how the machine is configured for your component, your contaminant, and your production flow.
This is where a consulting approach makes the difference: analyzing parts and dirt, defining the cycle, choosing technical solutions (filtration, handling, nozzles, parameters), and testing until a stable result is achieved. Once the process is stable, it also becomes measurable: same inputs → same outputs → simple and reliable controls.
If you're working to take your cleaning from visually clean to demonstrably clean, you can start with Geicos solutions: automatic, manual, and ultrasonic parts washers, which can also be customized to your needs.