How AI and Science Are Helping Make the World’s Food Safe
Originally published on LinkedIn by Abigail Stevenson, Vice President, Mars Advanced Research Institute at Mars.
The number of people facing extreme hunger is expected to nearly double as a result of the COVID-19 pandemic—to a total of more than a quarter billion.
This is unacceptable.
At Mars, we know it can’t be food unless it’s safe. And as a major global food company, we believe we have a particular responsibility to make food safe throughout the world.
We also believe food safety is a fundamental part of food security—the right for people, at all times, to have access to sufficient, safe, nutritious food.
The COVID-19 pandemic has highlighted—and accelerated—the risks and fragility of food supply chains. It has also shown that food safety and security risks are constantly evolving, therefore requiring novel approaches in order to be properly addressed.
Throughout almost three decades at Mars, and now as the Vice President of the Mars Advanced Research Institute (MARI), I have always felt fortunate to be able to leverage my passion for science and technology for the benefit of society. I recently spoke at the International Association for Food Protection’s Asia-Pacific Food Safety International Conference on how MARI and Mars are using scientific discoveries and Artificial Intelligence (AI) to help ensure food safety and security for people and pets.
A shift in our approach to food safety management
Traditional food safety management follows a ‘detect and destroy’ approach: if a problem is detected through sampling and testing, the food—now considered unsafe—will be destroyed. This leads to wasted food, energy and resources.
At MARI, we’re collaborating with the Mars Global Food Safety Center (GFSC), other teams across the Mars network as well as global partners to combine our existing knowledge of food safety with emerging technologies such as AI. The aim is to develop innovative platforms that allow for a more proactive, prescriptive and preventative approach to food safety management based on a better understanding of the factors driving test results.
In other words, we’re shifting from ‘detect and destroy’ to ‘predict and prevent.’
What does this mean in practice?
Take aflatoxins for example—poisonous but little-known compounds made by certain fungi. Aflatoxins can be found in corn, wheat, peanuts, tree nuts, spices and other staples. Each year around 4.5 billion people are exposed to these deadly toxins, which are known to contaminate as much as 25% of the world’s food crop area. They are among the most potent naturally occurring chemical carcinogens known to humans—and can therefore have a staggering health impact if consumed by people and also pets.
Eradicating aflatoxins is therefore a priority for human and pet health and for the safety of our supply chains. Doing so requires an interdisciplinary approach to gain an understanding of the different factors influencing aflatoxin production in every step of the supply chain. This includes the genetics of the crop varieties on which it grows; the temperature and moisture during the harvesting, manufacturing and storage of these crops; the soil conditions in the field; and much more.
Data from a range of different disciplines is, therefore, needed in order to truly understand the risk posed by aflatoxins and identify potential solutions. The data is a common language for all the disciplines to communicate in—but enormous amounts of it need to be collected and analyzed. It’s a seemingly impossible task for scientists to do alone.
And that’s where AI comes in.
Using this transformational technology, it’s now possible for all these layers of data to be integrated and analyzed. We can therefore start predicting what factors will lead to aflatoxin production—and take steps to prevent this.
Aflatoxin risk management is just one of the ways in which MARI, together with the GFSC, is combining our knowledge of existing food safety practices with AI to develop new platforms and capabilities that help ensure that ingredients, processes, and ultimately, food products are safe.
The pandemic has highlighted the increasing complexity of food safety challenges. At Mars, we believe that interdisciplinary collaboration combined with the power of data and AI are crucial in tackling these. That’s why we’re making these future-focused approaches an important part of our commitment to making the world’s food safe—during the pandemic, and beyond.