Define the measurand
Agree the material, wet or dry basis, measurement volume and decision to be supported.
Quality and validation
Turn a sensor reading into traceable process evidence
A credible moisture project links every displayed value to a defined reference method, representative samples and documented acceptance conditions.
Quality and validation
A traceable result is built as a chain
Skipping one link can make a stable-looking value unsuitable for control, quality release or settlement.
Choose the reference
Use the applicable laboratory, oven or recognised industry method and record its conditions.
Plan sampling
Match online time and location while protecting samples from moisture loss or gain.
Build the dataset
Cover real moisture, density, temperature, formulation, particle size and seasonal variation.
Calibrate the model
Relate sensor response to reference results with a controlled model and version record.
Validate independently
Test data not used to build the model and compare results against pre-agreed criteria.
Verify in service
Monitor bias, process changes, maintenance and the triggers for revalidation or recalibration.
Quality and validation
Six workstreams behind calibration confidence
Each workstream answers a different risk. Together they define whether the result remains valid in the real process.
Reference method
Confirm method scope, sample preparation, drying or laboratory conditions, calculation basis and current revision.
Representative sampling
Cover space, time, depth, batch, supplier and process state; pair each sample with the correct online interval.
Dataset design
Include the full intended range and influential variables; separate model-building data from independent validation.
Performance evidence
Review bias, residuals, repeatability and errors by material group and operating range—not one headline number.
FAT and SAT
Document configuration, I/O, installation, communications, paired sampling, alarm logic and agreed acceptance tests.
Ongoing verification
Use routine comparison, trend review, change records and clear triggers for adjustment or full recalibration.
Quality and validation
What a complete validation package should contain
The package should allow another engineer or quality manager to understand what was tested, under which conditions and where the limits are.
- 01Reference-method agreement
- 02Sampling and dataset matrix
- 03Calibration and independent-validation report
- 04FAT/SAT and maintenance plan
Published performance must remain tied to the tested material and installation conditions.
All blog articles
Related engineering guidance
How to select a laboratory reference method for an online moisture system
The online system learns a relationship to a defined reference result. It cannot correct an unsuitable, inconsistent or poorly controlled laboratory method.
Representative sampling: connect the laboratory sample to the online value
A laboratory result is useful only when the sample represents the material volume and time associated with the online reading.
Design a calibration dataset that covers real production
A calibration built from convenient middle-range samples can look excellent and still fail at the operating limits that matter most.
Complete knowledge centre
Continue through the full moisture decision library
Compare measurement routes, sampling, calibration, validation and production-control guidance.
34professional articles
Content paths
Browse all articles - Online & laboratory
- Measurement technology
- Calibration and validation
- Production applications
Contact Us
Start validation planning before installation
Share the material, measurement point, reference method and intended decision so the sensor route, sampling access and acceptance plan can be designed as one system.
Tell us what you measure, where it moves and how the result will be used
Share a few process details and our application team will recommend a suitable technology, installation concept and validation plan.