Plastic identification for sorting

We have defined predictive models for the recognition of the main plastics. Using miniaturized sensors, we give operators the possibility to classify them for sorting operations. Our algorithms can also be used on hyperspectral cameras, for image analysis.

Methanization process control

Our solution can measure five parameters in digester inputs in real time and provides a predictive model to control the productivity of the process.

Meat defrosting process control

Using a miniature sensor communicating with a smartphone by Bluetooth, we measure food temperature while defrosting: the solution is adapted to the control of the cold chain.

Identification of melamine in milk

With miniaturized sensors, we are able to identify traces of melamine in powdered milk. We use PLS quantification models, adapted preprocessing. We obtain an RMSE of 3% in weight.

Identification of nutritional value

In fruits and vegetables, we measure nutritional values ​​with our predictive models: sugars, proteins. Our models are also suitable for the control of the varieties.

Identification of stains on textile

We have developed models able to identify spots on textiles and characterize them using UV-VIS-NIR cameras. We can help you improve your cleaning processes and your brand image.

Biological origin of textiles

Pour certaines matières, nous avons établi des modèles prédictifs pour reconnaître une origine bio d’une origine non bio de la matière. Ces modèles peuvent vous aider à valider vos approvisionnements lors de vos opérations de contrôle d’entrée

Adultération of essential oil

For cosmetic applications, we have developed models to measure adulteration in essential oil. We can distinguish lemon eucalyptus oil from citronella oil, for example. This process guarantees the quality of the materials used. Tell us about the measure of adulteration you want to control.