Bühler innovates with Microsoft to tackle toxic grain contamination

Food technology company Bühler has announced a breakthrough in sorting technology, called LumoVision, that will mitigate toxic contamination in maize, as well as improving yields by identifying and removing aflatoxin-infected grains, known to cause cancer.

The company has partnered with Microsoft to develop LumoVision, described as a ‘significant advancement’ for the grains industry, defending against aflatoxin and other mycotoxins.  The sorting technology is reportedly faster and more efficient than existing solutions and can eliminate up to 90% of contaminated maize.

Bühler managing director of digital technologies Matt Kelly said: “Advances in digital technology, together with our sorting and food safety expertise, make this an unrivalled system that contributes to solving a major global food safety and security challenge.”

Bühler food safety officer Beatrice Conde-Petit said: “Aflatoxin is one of the biggest global pains today. Climate change means that the problem is growing, so eliminating contaminated grains from the food chain more effectively, with less loss of healthy grain, is an urgent challenge.”

Maize is considered an important global food crop, consumed by 500 million people, and is used as a major ingredient in animal feed.

It is said that two contaminated maize kernels in 10,000 is enough to spoil a whole yield. LumoVision is the first optical sorting technology able to identify toxins based on direct indicators of contamination. It uses real-time cloud-based data to analyse risks of contamination.

Previous methods such as thin-layer chromatography, high-performance liquid chromatography, mass spectroscopy, and enzyme-linked immune-sorbant assay have proved inconclusive and time-consuming, relying on indirect indications.

The technology analyses the colour of each kernel under fluorescent UV lighting and identifies kernels that glow a specific bright green colour as contaminated. LumoVision is equipped with highly sensitive cameras to detect this exact shade.

Air nozzles then blow the contaminated kernels out of the production line within milliseconds of detection. The system can process up to 15t of kernels per hour.

Microsoft’s cloud-based infrastructure then helps to reduce overall yield loss by combining the data from the cameras with existing information stored in the cloud to provide a risk analysis. The system is known to reduce yield loss to below 5%, compared to current solutions operating at 5%-25% yield loss.

Bühler senior research engineer for digital technologies Ben Deefholts said: “We are incredibly excited about this achievement. As an organisation we have strived to solve the problem of aflatoxin contamination for many years. Now, with today’s technological advances we’re able to bring this ground-breaking solution to market.

“With data science techniques and tools we can develop sorting algorithms, while connectivity and internet-of-things (IoT) solutions allow us to combine our optical sorting with real-time risk models.”

LumoVision is being tested at Italian agricultural cooperative Capa Cologna, which was affected by drought conditions in 2012, leading to an increase in aflatoxin contaminated yields.