Dec. 11, 2019
Seventeen Moore School and College of Engineering and Computing students participated in the inaugural Analytics Hackathon in November to solve an operations and supply chain real-world issue for Schaeffler Group USA Inc. A global automotive and industrial supplier, Schaeffler Group USA Inc. approached the Moore School about creating an event where students could examine their warehouse’s inventory levels to boost efficiency, reduce holding costs and maximize cash flow.
The student teams for the event were tasked with analyzing real inventory data provided by Schaeffler Americas to identify potential improvements the company could make in lowering inventory levels while maintaining or improving service levels. Schaeffler Americas was not only interested in creative solutions USC students could provide, but the company was also looking for new ways to partner with universities, said Shoukat Ali Bhamani, chief information & digital officer for Schaeffler Americas.
“Schaeffler thinks the students from USC did an excellent job in the Hackathon,” said Frank Zhou, Schaeffler Americas data management manager. “We ranked the top two teams who both were able to deliver innovative concepts and prototypes, and we will work further to realize their solutions.”
The winning team included Kirk Ryan Fiedler, chemical engineering student with a minor in data science; Matthew Owens, an international business and operations and supply chain student with a minor in Chinese; Skylar Ramsay, a marketing student with a concentration in business analytics; and Alexandra Sakellar, a finance student with a minor in applied computing. The winning team won a $1,000 total prize; the event was sponsored by Schaeffler Americas and BetaBox Learning, an education technology company that augments K-12 technology education programs based in Raleigh, North Carolina.
The teams received the data a week before the Hackathon and worked with Schaeffler Americas and BetaBox Learning executives and Moore School faculty and Data Lab staff who are well-versed in data analytics during a six-hour power-session. The next day, the teams presented their findings to a panel of judges.
“The Hackathon event provided students with an experiential learning activity that engaged their creativity,” said Linda Oldham, director of Corporate Partner Engagement for the Operations and Supply Chain Center and the Center for Applied Business Analytics at the Moore School. “This was an excellent opportunity for students to hone their analytics, problem-solving and teamwork skills.”
The winning team took a different approach from the other three teams; many of the other teams did some type of linear regression model to give Schaeffler Americas an actual number value, Owens said.
“We decided to create something a bit different. We identified situations where the most amount of optimization potential is present and then pulled out two examples to show them how they could implement our changes,” he said. Team member Fiedler “greatly improved upon our team's process by creating a neural network model which could be used to predict trends in inventory levels with only a 10 percent error margin.”
Fiedler said the trend analysis that resulted from his neural network model allowed for better control over the inventory levels. Using real-world data for the project, Fiedler said he learned to be adaptable when working with a complex data set.
“Going into the competition, I had all of these different ideas for analysis,” he said. “Upon receiving the actual data set, though, it immediately became clear that I would not be able to use these tools of analysis and would need to come up with a method that would work within the data I had access to. I had a great time devising and implementing our strategy.”
Sakellar said the Analytics Hackathon was an amazing learning experience, from drafting their team’s strategy to presenting their findings to the Center for Applied Business Analytics advisory board.
“The Analytics Hackathon tested my knowledge and ability to apply what I’ve been learning in my classes,” she said. “I had a lot of fun working on the data visualization process and received awesome guidance from the MBA students [Jerryl Don Lee and Michael Morse during the power session]. The event introduced me to new concepts and allowed us to view the project from multiple perspectives.”
All of the students who participated in the Analytics Hackathon were able to showcase the skills they’ve learned in their courses with a real-world application, which is a practice the management science department highly recommends, Jack Jensen said. Jensen is the co-director of the Operations and Supply Chain Center in the Moore School and a clinical professor of operations and supply chain management.
“This Hackathon required students to apply their domain specific knowledge of inventory theory and inventory modeling coupled with their ability to employ analytics tools that are capable of digesting vast data sets,” he said. “What we saw in each teams’ final products were very creative and useful ways to identify the most problematic inventory issues as well as a solid path forward to address these issues.”