A Google affiliate has created two new programs that make it easier for food producers, suppliers and commercial kitchens to send unnecessary food to food banks that need it.
Known as Project Delta, the machine learning programs take into account thousands of different calculations and variables, the things that a dedicated team of organizers would need to manage, to ensure that food goes where it is needed most, where it is most likely to arrive. eaten, and other priorities.
Food waste is what sportscasters would describe as “a good problem” since it inherently suggests that there is enough for everyone. The problem is, he doesn’t always go where there are hungry people.
The problems are mainly in the supply chain: through the interaction of buyers, trying to imagine how many units they need, and sellers trying to imagine how many they can sell.
Producers, like sellers, try to earn as much as they think buyers can find, while the ultimate food owner (supermarkets, restaurants, or hotels) often has too many processes to worry about to consider how best to send food further. down the line.
Efficient food distribution is an extremely difficult job to do throughout the supply chain, so it is perhaps not surprising that 30-40% of food in the US goes to waste.
“There is no simple way for food providers to inform food banks about what they have available, or for food banks or pantries to communicate what they need,” Adele Peters writes to Fast company, covering Google innovation.
The idea is being worked out at a Google branch called X, which calls itself “The Moonshot Factory” and is focused on providing super innovative solutions to make the world a better place.
A better way to bank
Emily Ma, writing for X’s blog On the two years of development and testing of Project Delta, he explains that they set out to “create a smarter food system, one that knows where the food is, what state it is in, and where it is best directed to ensure it doesn’t end up in a landfill. and instead it goes to the people who need it most. “
Some of the issues that those working on the issue at Google faced were things like the lack of an industry or cross-industry standard on how food providers communicate what they have or need to move the most, or in terms of banks. of food, what they need. and what people don’t eat.
She jokes that there is not even an industry standard designation for the state of Texas, and that during her and her team’s preliminary investigation, they found 27 different words in organizational data, i.e. TX, Texas, Tx, etc.
Working with Southwest Produce Cooperative (SWP) food banks in states like Arizona, Ma and his team first built their machine learning prototype, which instead of phone calls, emails, site visits, and paper records, forms Common where SWP food bankers coordinate shipments – Uploads all supply and demand related logs to the algorithmic bot detailing what needs to go, where and when.
Then Emily Ma and the Moonshot Factory team went to Kroger to see if they could improve logistics at Feeding America, the nation’s largest national hunger relief organization.
Their program’s insights into the operation of Kroger delicatessens, where meat and other products are often recycled due to health and safety concerns, enabled them to open opportunities to provide millions of additional meals to communities in need. .
Finally, in a more impressive display of machine learning, cameras installed next to trash cans in Google’s facility kitchens were able to collect twice as much information on food waste as manual records made by chefs who took between 30 and 60 minutes. to complete.
Machine learning can identify trends, such as wasting large amounts of a particular food, as well as make recommendations for dishes and ways to recycle that can reduce food waste in commercial kitchens.
SHARE this fascinating story with friends on social media …