OpenScienceOpenScience
PROPOSE A CHALLENGE PT
PRODUCTS

In-house equipment and software

Custom hardware and software for your needs.

Olive oil study with FT-NIR

Three charts: raw FT-NIR spectra of 36 spectra from 12 olive oil samples, the 7000 to 9000 cm⁻¹ region after SNV, and the principal component analysis

What the data are

36 FT-NIR spectra from 12 olive oil samples, with 3 replicates per sample, acquired from 12500 to 4000 cm⁻¹. Each color is one sample.

What the charts show

On the left, the raw spectrum. In the center, the 7000 to 9000 cm⁻¹ region after SNV, a standard preprocessing that removes intensity differences between spectra. On the right, the principal component analysis of that region: the replicates of each sample stay close together, and the samples occupy different regions of the plane.

How to read it

Replicates close to each other indicate the measurement is reproducible. Samples at different points of the plane indicate the spectrum carries differences between them.

OS4CI with the Imajan HS4017

OS4CI result screen: sample composition in coffee, picking, impurity, large, moka, moisture and cup, class table, bean mosaic by class and camera status

Sample result

Coffee, picking and impurity by the Brazilian Official Classification, large and moka by sieve, and moisture and cup estimates, with the interval of each value and the count per class.

Beans and machine

Mosaic of the actual beans by class, with the probability of each, and the status of the visible and infrared cameras: temperature, frame rate and dropped frames.

Ready to issue

Before releasing the result, the interface lists pending items and warnings, and every step is recorded in the sample log.

Review of ambiguous beans: the classifier shows the most likely classes of each bean for the analyst to decide

Bean review screen with ambiguous beans at the top and probabilities per class

Sample color: projection of the beans in CIELAB space

Screen with the projection of bean color on the a* b* plane of CIELAB space
RESEARCH AND DEVELOPMENT

What we are building now

In-house R&D fronts. Interested companies and cooperatives can take part in the field validations.

Mycotoxins and pesticides

Multispectral reader for rapid tests

Turns commercial immunochromatographic tests (lateral flow) into a quantitative, traceable result, right at grain receiving or storage.

  • QuEChERS sample preparation adapted for use outside the lab
  • Multispectral imaging system
  • Onboard logistic calibration (4PL and 5PL) and chemometrics
Proof of concept completed. Scaling up to an industrial prototype and field validation.
Volatile compounds

Electronic nose

Array of gas sensors with an ambient sensor and a controlled sampling pump, to recognize and classify volatile compound profiles.

  • Product aroma and freshness
  • Fermentation, spoilage and contamination
  • Quick screening before chromatography

Custom electronics and classification models trained on the client's samples.

Hardware in development.Board in layout phase.
Chromatography

GC-MS: deconvolution and deformulation

Separates coeluting peaks, aligns runs and identifies compounds by spectral similarity against reference libraries, on the OS4DC platform.

  • Curve resolution by SIMPLISMA and MCR-ALS
  • Deformulation: finds the mixture of references that reconstructs the sample
  • Fuel-marker fraud in spirits
  • Fragrances and aromas with high peak density
In validation with real data from an official laboratory.

Rice in three bands, before and after extended focus

Rice grains at 550 nm in a single plane, partly out of focus
550 nm · single plane
The same rice grains at 550 nm with extended focus
550 nm · extended focus
Rice grains at 700 nm in a single plane, partly out of focus
700 nm · single plane
The same rice grains at 700 nm with extended focus
700 nm · extended focus
Rice grains at 800 nm in a single plane, partly out of focus
800 nm · single plane
The same rice grains at 800 nm with extended focus
800 nm · extended focus

The same field in three bands: the image changes from one band to another, and that difference feeds the chemometrics. With extended focus, each pixel’s spectrum uses only the bands in focus. Multispectral microscopy images of rice grains.

Want to test it with your own samples?

Remote or in-person demo in Campinas, Brazil. The software can be evaluated with your own data.

REQUEST A DEMO