Carnegie Mellon University

food

June 21, 2018

MBA Students Partner with Pittsburgh Nonprofit to Address Food Waste

“If you really want to get good at something,” said Ian Roddy, MBA ’18, “you need to actually do it, and do it in the real world.”

Roddy was part of the Business Analytics MBA track capstone course in the Tepper School of Business this spring, in which he and MBA classmates Natasha Kosik, Alec Burtner-Abt, Brian Stump and Ian Very worked with 412 Food Rescue to address a real business problem the nonprofit was facing.

In addition to coursework, Tepper MBA students can apply to participate in one of five tracks focused on a particular field of study or career path. The Business Analytics track, run by Willem-Jan van Hoeve, Carnegie Bosch Associate Professor of Operations Research, focuses on understanding and interpreting business data to guide decision-making. Students in the track must take three courses during their MBA studies and complete a capstone project that challenges students to put into practice the skills and knowledge attained in their coursework.

“There’s kind of a gold rush for analytics talent,” Roddy said. “I wanted to position myself well for a career.”

Van Hoeve approached 412 Food Rescue’s founder Leah Lizarondo, HNZ ’03, about participating in this year’s capstone course. The organization has worked with student groups at Carnegie Mellon University and other universities previously, “because we tackle a really interesting problem,” Lizarondo said. “We waste 40 percent of our food, and one in seven people are hungry. We create the infrastructure and services that will enable food recovery in urban regions.”

Lizarondo gave the student group the opportunity to select what problem they wanted to address. “When I work with students, I really want them to work on what they’re interested in,” she said. “When they’re inspired by a problem, they can work on it better.”

“They basically told us, ‘Here’s what we do. How do you think you can help?’ They put it on us to tell them what we wanted to work on,” Roddy said, “which we really appreciated.”

The students visited 412 Food Rescue’s headquarters and observed its operations. “What I was most impressed with was how they really invested the time to learn what our business was,” Lizarondo said.

The organization works with nearly 400 retail locations in the Pittsburgh area — including Giant Eagle, Whole Foods and Trader Joe’s stores — to collect food that might otherwise be discarded. After collecting the food, they distribute the donations to more than 100 nonprofit organizations who serve food-insecure Pittsburgh-area families.

Roddy reported that the organization was concerned that it was not equitably allocating the donations under the current process. “Everyone had their own set of factors that they were considering when deciding where to send a donation. They didn’t have a very consistent set of criteria to help them decide where to send stuff.”

412 Food Rescue processes between 18,000 and 20,000 transactions each year, Lizarondo said. “When you’re only matching a few, the human brain can pretty much comprehend how to do that. But as soon as you’re matching tens of thousands of food sources, the process starts to become more inefficient. And so there really needs to be some intervention that happens to prevent that.”

In addition to dealing with scale, one of the challenges Roddy highlighted was a timing factor: Not only was the pickup time restricted to the operating hours of the retail location, but the food — most of which is fresh produce — must be delivered to the recipient organization as soon as possible, so that the end consumer receives a fresher product with a longer shelf life.

The organization had much of its data spread across multiple systems, so Roddy’s team decided to develop a single user interface that could help employees make efficient and equitable choices about how to distribute food donations.

Roddy noted that 412 Food Rescue had a license for Tableau, software designed for data analytics and visualization. He had basic knowledge of the software and of the programming language SQL from a Modern Data Analytics course he had taken at the Tepper School.

Together with Kosik, Roddy developed a “deservingness score” that Tableau would calculate for recipient locations each time a donation was reported. The score factored the distance between the donor and recipient locations, the amount of time since the last donation to that recipient, and the volume of donations collected by recipient locations. Tableau would then plot the locations on a map with a color code — redder locations had a higher deservingness score.

Although the students were working with dummy data for this project, Lizarondo reported that 412 Food Rescue is in the process of connecting the Tableau interface to the organization’s real database. “It will form the logic behind the basic matching algorithms that will eventually be integrated into our technology platform.”

Roddy said, “We all felt really good that the work we did is going to get used, and they appreciated it.”

Following his recent graduation from the Tepper MBA program, Roddy is joining the Ford Motor Company’s planning and strategy group. “I expect to use all the skills I learned in this capstone,” he said. “The experience of taking a really wide-open problem and applying some structure to it — that’s going to be invaluable in my career, I think.”