Harvard Dining Services Expands Artificial Intelligence System to Prevent Food Waste

Harvard University Dining Services (HUDS) has fully implemented the AI-powered Winnow system across all residential dining halls and Annenberg to track and reduce food waste.

Harvard University Dining Services has activated the Winnow system in all dining halls, having begun campus-wide rollout last spring following a pilot program last fall. According to a report shared by the company, the system helped prevent 23,500 pounds of food waste as of September. Using a scale and overhead camera to weigh and photograph food scraps, the system uses HUDS's weekly menu data to identify items and produce reports listing the dining halls' most wasted items by volume and cost.

Winnow Development Manager Gina Sears stated that these reports provide each dining hall management with the opportunity to see the menu items generating the most waste. As part of the measures taken, chefs noticed an average of 11.5 pounds of eggs being wasted per day and began preparing eggs in smaller portions, dropping daily egg waste to 3.1 pounds. Additionally, the menu cycle was reduced from four weeks to three weeks, and the average food waste per meal decreased from 2.12 ounces to 1.46 ounces.