At KorNutra, we see near-infrared spectroscopy and texture analysis as powerful tools for objective, repeatable quality data. But if you tried to automate the entire sensory quality check with only those two instruments, you would lose several layers of sensory information that remain uniquely human.
What the instruments capture
Near-infrared spectroscopy gives a chemical fingerprint and can flag gross compositional changes, moisture differences, or blend inconsistencies. A texture analyzer measures mechanical properties such as hardness, springiness, cohesiveness, and fracturability. These are useful, but they only cover part of the sensory picture.
What a human tester can still detect
- Aroma and volatile character. NIR may detect broad chemical changes, but it does not smell the product. A trained human panel can identify rancid, oxidized, grassy, earthy, fermented, or solvent-like off-notes at levels that are difficult to calibrate spectroscopically.
- Taste and flavor balance. Sweetness, bitterness, sourness, saltiness, umami, and botanical flavor notes are perceived in combination. NIR predicts concentration, not perceived intensity or balance. It cannot tell you whether a batch tastes right or has a subtle off-flavor.
- Trigeminal and mouthfeel effects. Cooling, warming, tingling, astringency, numbing, and throat catch are sensory events, not simple texture measurements. A texture analyzer measures force, but it cannot replicate how a powder, tablet, capsule, or gummy feels as it dissolves or breaks down in the mouth.
- Dynamic oral processing. Humans perceive how a product changes over time: initial hardness, chew-down, particle breakdown, coating release, residue, and aftertaste. A texture analyzer provides a single force profile, not the full time-intensity experience with saliva and oral temperature.
- Visual and macroscopic defects. NIR gives spectral data, but it does not see the product the way an inspector does: speckling, mottling, gloss, edge chipping, color uniformity, clumping, stratification, surface pitting, or foreign matter. These visual cues are often the first sign of a process deviation.
- Overall sensory integration. A human panel can judge whether the product meets the KorNutra sensory target as a whole, including aroma, taste, texture, appearance, and aftertaste together. Instruments measure individual attributes but cannot make that integrated judgment.
Calibration still needs the human panel
The instruments still depend on people. NIR does not measure sweetness or rancidity directly. It records near-infrared absorbance, and a chemometric model converts that spectrum into a predicted value using reference data from lab analyses or panel scores. Texture analyzer limits come from the same kind of reference set: manufacturers set a pass/fail threshold from the range of batches a trained panel accepted. When the formula, flavor, or supplier changes, you have to rebuild the model and those limits against fresh human judgments. Skip that step and the instrument can pass an off-flavor lot because its reference set never included that off-note. On an instrumented line, the panel's main job becomes defining what the instruments should flag.
At KorNutra, we use NIR and texture analysis to reduce subjectivity and catch deviations early, and we keep trained human sensory testers in the loop for exactly these reasons. The goal is to use automation where it is strong and keep human sensory evaluation where it remains essential.