Message from Mars: Big Data and Genomics Can Make Our Food Safer
Salmonella. E. Coli. Listeria. Killer bacteria are wreaking havoc on society. Nearly 1 in 6 people contract a foodborne illness every year. In the US alone, thousands die. The economic toll is a staggering $77 billion. And if it seems like things are getting worse, blame our global food supply—and the status quo process of bacteria detection. But there’s hope on the horizon. A wild-eyed scientist at Mars is working with IBM to create a radical new system to stop pathogenic spread. It combines big data and genomics with the latest thinking in medicine. They call it sequencing the food supply chain. If you care about your health, you’ll want to listen closely.
One of the world’s largest food manufacturers, Mars works with thousands of suppliers and millions of tons of raw materials every year. At the Petcare division in Nevada, where Pedigree and Whiskas are made, supplies come in primarily by rail.
Mars Petcare, and the food industry at large, has used the same method to prevent food contamination for decades. It relies heavily on supplier relationships, materials inspection, biological sampling, high-temperature cooking, and sanitation. By contrast, Mars Petcare’s pilot project uses the latest tools and methods borrowed from genomics, big data and modern medicine.
Why start with pet food when human lives are at stake? “We’re using the bits that people don’t want. It actually gives us a bigger challenge from a safety point of view. It’s harder to do,” says Mars VP of corporate R&D Dave Crean.
A tour of Mars Petcare reveals a highly automated operation. From the storage silos through final packaging, materials encounter minimal human interaction. By reducing handling, Mars reduces the risk of contamination.
North American director of quality and food safety Annette Miller leads a tour of the facility. It smells like you might expect—but is probably far cleaner than you imagine. Machinery is disassembled and scrubbed every weekend.
Factory bosses like Reno site manager Andrew Everitt (back) could benefit greatly from Crean’s new system. Today’s sampling methods are slow and scattershot. A “factory microbiome” approach will be far more encompassing and immediate, triggering an alarm whenever a foreign substance finds its way into the factory.