X, the "moonshot factory" of Google parent company Alphabet, stated that it had developed two prototypes as part of a project called Project Delta have graduated and are now being transferred to Google for scaling and commercialization. The company's mission was to create a smarter food system that knows where the food is, what state it's in, and where best to direct it to ensure it doesn't end up in a landfill and instead goes to the people who need it most.
The first is an intelligent food distribution network," nicknamed "dana-bot." To build it, the X team took a dataset donated by the Southwest Produce Cooperative, categorized and standardized each entry, and then used it to match food in food banks and pantries based on "real-time needs in the Feeding America network. It has taken two and a half years for the company to prototype and tests a range of technologies to reduce food waste and food insecurity.
The second prototype utilized computer vision and machine learning to capture images of food thrown out in Alphabet kitchens. After running it in 20 different units across Alphabet cafes over six months, it was able to automatically collect information about the kitchen's food wastes as the manual system. The team plans to roll out its computer vision system to more Alphabet kitchens and utilize Google's resources to expand the food distribution network and eventually offer it to other organizations.