The question is almost never “can this technology measure that”. It is “at which step of your process does that measurement change a decision”.
PAT-INDUSTRY designs and deploys non-destructive in-line measurement. The measurement consumes no product: everything manufactured stays saleable. The table below reads by industry: for each unit operation, what is commonly measured there, and the page that details the technologies, the conditions for success and the limits.
Pharma
| Unit operation | What is measured there |
|---|---|
| Material receipt | Identity and conformity — see |
| Blending | Blend homogeneity — see |
| Granulation | Particle size and water content — see |
| Drying | Water content, and end of operation — see |
| Compression | Dose uniformity — see |
| Coating | Coating thickness and uniformity — see |
| Solubilisation and dissolution | Active and excipient concentration — see |
| Emulsification and homogenisation | Droplet size — see — and homogeneity — see |
| Viscous formulation | Viscosity, during the operation — see |
| Final control | Internal structure, through the pack — see — and identity — see |
Drug substance
Biopharma and biotech
| Unit operation | What is measured there |
|---|---|
| Cell culture | Glucose, lactate, density — see |
| Purification | Composition and identity — see |
| Formulation | Composition and viscosity — see |
| Lyophilisation | Residual water content — see |
| Medium preparation | Component identity and titre after reconstitution — see |
| Dispersed forms | (Nano)particle size and aggregation — see |
| Filling | Identity through the vial, by NIR and RAMAN — see |
Chemicals and polymers
| Unit operation | What is measured there |
|---|---|
| Reaction | Progress, by-products — see |
| Liquid formulation | Composition — see |
| Drying | Water content — see |
| Dispersion | Phase distribution — see |
| Crystallisation and wet milling | Size distribution — see |
| Identity checking and sorting | Polymer type, fillers and additives — see |
| Packing | In-line viscosity — see |
Cosmetics
Agriculture and food
| Unit operation | What is measured there |
|---|---|
| Receipt | Identity, water content, protein — see |
| Drying | Water content, and end of operation — see |
| Blending | Homogeneity — see |
| Fermentation and digestion | Medium composition — see |
| Flow in a vessel or a pipe | Phase distribution, and the product changeover front on a shared line — see |
| Finished product control | Composition and colour, in the visible and near infrared — see — counting and morphology — see |
Anaerobic digestion
| Unit operation | What is measured there |
|---|---|
| Feedstock reception and characterisation | Dry matter, organic matter, methane potential — see |
| Silage taken from the working face | Dry matter, from one day to the next — see |
| Running the digester | FOS/TAC, volatile fatty acids, ammonium — see |
| Digestate on the way out | Dry matter, organic matter, nitrogen — see |
| Biomass drying | Water content, and end of operation — see |
This table does not close the list
The operations and attributes above are the ones most often instrumented. They do not bound what can be measured. Part of our work is developing a measurement where none existed: an unusual matrix, an attribute nobody had tied to a physical quantity, a mechanical access thought impossible.
If your need is not in the table, it is not excluded from it: it has simply not been filed yet.
Two questions to ask before choosing a technology
Are you measuring a quantity or a state? A content, a concentration, a thickness are quantities: they need a model, therefore a campaign of reference samples and upkeep over time. A state — stabilised, homogeneous, conforming — is often obtained without a model. The two do not cost the same.
What mass of product does the measurement cover? A probe only sees the volume it illuminates. Depending on the technology that ranges from a few milligrams to several hundred grams, and it decides representativeness long before instrument performance does. What a sensor really sees.
Describe the operation, not the instrument
Forty-five minutes are enough to say whether a non-destructive measurement answers your question, and which one deserves a trial. Tell us your constraints too: the equipment already qualified, the classified area, the reference method you inherit, the pace at which the decision has to be made. Feasibility is often decided there rather than in the choice of technology.