Deep learning tests for contaminants in food factories
Spotted: Almost all of the food that we eat is processed in some way in a factory setting. These factories need to be kept very clean to avoid bacterial contamination, which can be an expensive and time-consuming process, involving constant monitoring and testing. French startup Spore.Bio has developed a way to speed up this process without compromising on safety.
Spore.Bio’s pathogen-detection system works by using a laser to shine an optical light of a particular wavelength on surfaces. Bacteria then react to this excitation in specific ways and the company trains its computer vision and chemometrics models to recognise this spectral signature, identifying the presence of bacteria.
To train the system, the light was first shined on a huge variety of surfaces, some that held clean food, and some that held contaminated food. The images produced were analysed by machine learning models that compared the two datasets to learn how to recognise and detect the presence of bacteria on a surface.
The company has recently completed an €8 million pre-seed funding round led by London’s LocalGlobe VC, with participation from EmergingTech Ventures, No Label Ventures, and several others. The funding will be used to further develop the technology.
Combatting food pathogens is the focus of a number of recent innovations spotted by Springwise, including a spray that kills harmful bacteria on food and a technique that helps plants combat fungal pathogens by disrupting the pathogen’s ability to cause disease.
Written By: Lisa Magloff