Higher Education’s Technology Shift Serves Up Food Service Cost Optimization

“We take a two-pronged approach to AI on campus,” shares José Rodriguez, vice president and CIO at Pomona College. “Administrative staff use Microsoft 365 and its AI-powered assistant Copilot for their work, while faculty have access to environments that allow them to build custom LLMs for the classroom and research.”

How Pomona College’s AI Solution Reduces Food Waste

Each dining hall kitchen has a smart scanning station that pairs a tablet interface with a 3D computer vision camera, weight sensors and an infrared sensor to capture food volume, weight and temperature in real time. The software that manages the system uses AI models that run both on the device and in the cloud.

“The edge computing piece allows for instant identification of food items and provides real-time feedback,” says Jose Martinez, general manager of dining services at Pomona College. “The cloud platform aggregates data into a dashboard, reports and forecasting tools. Altogether, it gives us a live, end-to-end view of what is being produced, consumed and wasted.”

With the solution in place, Dining Services has gained a level of visibility and control that staff did not have before, allowing them to make decisions in real time, rather than reacting after the fact. 

“We now have precise data on overproduction, consumption patterns and plate waste across different dining halls and menu items,” Martinez says. “For example, we have identified that Caesar salad, yogurt and Mexican rice are among our top-consumed items. At the same time, our data showed consistent overproduction of items like Mexican rice, pollo asado and broccoli. That level of visibility allows us to make targeted adjustments instead of broad changes.”

DISCOVER: Observability tools give IT teams better visibility into their networks.

Pomona’s AI solution has yielded tangible sustainability results, reducing overproduction by 54%. This also translates into real financial savings. Dining Services has seen an 8% reduction in plate cost, driven by better alignment between production and actual consumption. The solution has also improved consistency across locations, strengthened accountability and given staff clearer feedback on the impact of their work. 

“Food waste is a major component of our sustainability strategy, both from an environmental and operational standpoint,” Martinez says. “Metafoodx gives us the ability to measure and report on waste with a high degree of accuracy, which is critical for accountability. From a strategic standpoint, it has become a core tool for managing both cost and sustainability.”

While AI is being applied to some interesting use cases across campus, it has not disrupted this small liberal arts college’s focus on student learning. 

“The college’s infrastructure was already well positioned to support AI processing, but we have also been in the process of updating our fiber architecture topology across campus, as well as installing a GPU processing node on our high-performance computer specifically designed for AI calculations,” Rodriguez says. “Security is also a critical part of our response to new technologies, and we continue to ensure a safe computing environment through network segmentation design projects and high endpoint security standards to protect the most sensitive campus data.”

In addition, Rodriguez adds, the college has updated master services and data processing agreements with vendor partners to ensure contract language includes AI considerations where applicable.

“Customer engagement with the industry is an important part of keeping us all accountable in a new world of AI risks and opportunities,” he says.

AI Solutions Bring Technology to Campus Sustainability Efforts

Like many higher education institutions, Pomona College has an administrative office devoted to sustainability. These offices play valuable roles on campus, organizing campuswide resource conservation, promoting efforts among students and to prospective students, and integrating sustainability into curriculum and research. All of these efforts help align campus missions more effectively with the kind of real-world challenges that students will face after graduation.

“A lot of campuses have sustainability as part of their mission and their strategic plan, along with directors of sustainability,” explains Meghan Fay Zahniser, executive director of the Association for the Advancement of Sustainability in Higher Education. “Some of these roles are more closely aligned with academics, and others are aligned to operations. Food services and dining halls, in general, have some of the strongest roles in campus sustainability efforts.”

LEARN MORE: Smart buildings help institutions meet sustainability goals.

AASHE manages the Sustainability Tracking Assessment and Rating System (STARS), a transparent, self-reporting tool for institutions to measure and benchmark their sustainability performance, including food recovery and dining service procurement. Common sustainability strategies for dining services include implementing trayless dining, eliminating bottled water, reducing single-use plastics and prioritizing local food resources. 

Fay Zahniser believes that AI and other technologies will play a growing role in helping support campus sustainability efforts, including dining services moving toward local food procurement.

“Many colleges are rethinking how they procure and prioritize local food while balancing existing relationships with major food service providers,” she says. “That transition is incredibly complex, with institutions having to manage sourcing across a wide range of foods and individual suppliers. I’m hopeful that AI can play a role in helping colleges develop sourcing plans more quickly and efficiently.”

Georgia Tech Gains Management and Operational Benefits With AI

Georgia Institute of Technology, located in Atlanta, is another university taking advantage of AI-powered technologies in its dining services to help cut greenhouse gas emissions. Tech Dining implemented a food waste solution from Leanpath in its kitchens, which has helped reduce food waste by 12% since being implemented.

The AI-powered solution includes a computer vision camera to identify food, a digital scale to weigh it and a tablet interface used by staff to input the data. Collected data can then be accessed, reviewed and queried through a mobile app or a cloud-based dashboard.

“Leanpath serves as the eyes and ears of the chef and sous chef when they aren’t onsite,” says Houston Freeman, director of customer experience for Tech Dining. “It has made us more efficient, helping us understand more quickly what’s going on across our locations, both from an analytics and an operational perspective.”

READ MORE: Data analytics offer insights into university operations.

Being able to access collected data and use the tool’s built-in AI along with the dining services’ food management system has led to tangible kitchen management benefits. If particular foods are consistently thrown out, Freeman and his team will consider pulling them from the menu or adjusting the recipes. Another area is food inventory.

“If something doesn’t get correctly rotated, we have to throw it out because it went bad on the shelf,” Freeman says. “This gets tracked in the Leanpath system. We can then have conversations with the store clerks to figure out why it didn’t get rotated properly or didn’t get used before it went bad.”

On the operations side, the solution has yielded upskilling opportunities for kitchen staff. Data analysis revealed that too much trim waste from salmon and fruit preparation was ending up in the garbage. The cause of this excess trim waste was determined to be inefficient knife skills by the staff. 

“It helped us pinpoint some opportunities for staff training improvements so we could address them,” says Freeman. “One of the things we enacted from that insight was proper knife training. We now do an annual training with all of the cooks to show them how to properly trim these different foods and reduce waste.”

The Ohio State University Uses AI To Integrate Dining Services Data

With a student population of more than 60,000, Dining Services at The Ohio State University, located in Columbus, Ohio, is responsible for a lot of food production. Given the university’s scale, its sustainability goals, which include diverting food waste from landfills, have real impact.

Dining Services at OSU recently deployed a data capture system across its kitchens. StreamLine consists of a computer vision-enabled smart scale that captures both the weight and image of logged food items. Data is collected in Topanga’s cloud platform, which can be queried through Orbit, StreamLine’s AI assistant.

“We were looking for a more structured and consistent way to capture data at the point of production without creating additional administrative burdens for operators,” shares Rahul Ponappa, associate director of operations for Dining Services at OSU. “Any tool we considered had to work within a live service environment where teams are focused first and foremost on feeding students efficiently and maintaining food quality.”

Dining Services has StreamLine connected to the university’s kitchen and menu management software. This data connection helps Dining Services better integrate its operations data with its production, menu management and waste data.

“This connection is important because it creates a feedback loop,” Ponappa says. “What we produce, what we serve, what is left over and what is wasted can all inform future menu planning, forecasting, purchasing and production decisions. Over time, that gives our teams a clearer view of what is happening at the station level and how that compares with what was planned.”

The biggest benefit OSU Dining Services is gaining from StreamLine is increased visibility, allowing it to shorten the time between identifying an issue and responding to it. 

“Historically, a lot of food waste conversations in food services have been based on periodic audits and anecdotal observations,” Ponappa says. “Having more consistent production and waste data allows teams to spot patterns earlier and make more informed adjustments around forecasting, batch cooking, menu planning and production levels.”