The University of Massachusetts Amherst is making significant strides in waste management and sustainability through partnerships with two AI-based companies, MetaFoodX and rStream. These collaborations aim to tackle the persistent issue of contaminated waste on campus, ultimately saving the university over $1 million in dining and waste management processes.
MetaFoodX, an AI food operations company, is helping UMass minimize food waste before it even reaches the dining halls. Their AI scanners weigh trays of food before and after students eat, tracking consumption patterns and high-demand periods. This data allows UMass Dining to adjust food production in real-time, reducing overproduction and saving costs. “After [MetaFoodX’s] system has been in place for enough time and covers enough of the service, we can start applying that information to our procurement, meaning we will order less food, knowing that we will have what we need,” said Kathy Wicks, UMass Dining Director of Sustainability. While exact savings are still uncertain, Wicks estimates that the university could save $1 million or more by optimizing food procurement.
Meanwhile, rStream, co-founded by UMass alumni Ian Goodine and Ethan Galko, is addressing waste contamination after students have eaten. Their AI system sorts contaminated waste into clean streams with minimal human involvement. “Nobody’s doing what we’re doing exactly,” Goodine said. The technology uses a camera and AI vision system to identify and categorize waste items, continuously updating its memory with each experience.
The impact of rStream’s technology is substantial. Ezra Small, the campus sustainability manager, noted that contamination rates on campus are over 50%. By correctly sorting recyclables, UMass can avoid paying higher disposal fees for trash. For instance, sorting 500 tons of recyclable material out of trash streams could save the university approximately $16,075. Additionally, the university could generate revenue by reusing sorted plastic waste to create recycled materials, such as clothing, to sell at the campus store.
Despite the promising potential, both technologies are still in the early stages of implementation. MetaFoodX’s scanners are being fully implemented across campus dining halls, while rStream’s technology is in the experimental phase, with limited availability and outreach. Goodine noted that it will take “one or two loops around the sun” before rStream has all the necessary information to begin engineering a permanent system.
The adoption of these AI technologies could revolutionize waste management not just at UMass but across the sector. By demonstrating the feasibility and benefits of AI-driven solutions, these initiatives could inspire other institutions to invest in similar technologies. Moreover, the focus on reducing waste generation and improving recycling practices aligns with broader sustainability goals, potentially influencing policy and practice in the water, sanitation, and drainage sector.
As UMass continues to refine and expand these partnerships, the long-term impact on waste management and sustainability practices remains to be seen. However, the initial results are promising, and the university’s efforts serve as a compelling case study for the potential of AI in transforming waste management practices.