For the past year and a half, the Studierendenwerk Vorderpfalz has been using delicious data's artificial intelligence in its merchandise management system at the Landau and Ludwigshafen sites - the result of the AI: 38 percent less food waste by avoiding overproduction that can no longer be used. While 74 meals had to be thrown away on average per day in the same period in 2018, the figure was only 46 in 2019.
Imagine: Exam week, Thursday is a holiday and bad weather with storms is forecast for Wednesday! What are the expectations for Wednesday's meal numbers? One of the biggest challenges for a chef in a commercial kitchen is to calculate exactly how many meals he needs to make that Wednesday. Many factors play a role in this, and only by looking at all of them will you have a chance of reducing the amount of food thrown away that is produced too much. Reproducing food is also not always easy in a large kitchen, and it is of course annoying for the guest if the lunch he wanted is suddenly no longer there.
In this planning process, the dining halls of the Studierendenwerk are supported by an algorithm. How it works: The cafeteria manager plans his meal numbers based on his experience. The algorithm, which is installed directly in the merchandise management system, checks the weather, events on campus, holidays, sales figures last time and other general conditions and suggests a forecast number of meals to the head chef. The chef can then decide whether to adjust his or her own forecast to match the algorithm's planning.
"In the first few months, the algorithm had to learn from us and continues to do so. The longer it is used, the more accurate its forecasts become and the more we in turn learn from it" Andreas Dubiel - Head of University Catering at the Studierendenwerk - elaborates on the collaboration with artificial intelligence.
"Less waste is not only good for the environment and our resources, but also helps keep food prices stable for as long as possible," he continues. After all, even food that is thrown away has to be paid for in purchasing. In the end, these costs would have to be passed on to our guests. A real win-win-win, therefore, if the garbage can remains empty!"
The software delicious data participated as a nominated applicant in the field of digitalization in the federal prize "Too good for the garbage can", which was awarded at the beginning of April 2019.
The use of "artificial intelligence" by delicious data" was one of four projects for which the Studierendenwerk Vorderpfalz received the Rhineland-Palatinate Environmental Award in 2019.
