The University of Maryland Data Challenge: How Booz Allen Is Mentoring the Future

How many materials targeted for recycling are actually being recycled? What kind of resources do we need to clean up our oceans, and how should these resources be allocated?

Answers to questions like these are important steps to a more sustainable planet, and over 180 undergraduate and graduate students were on the case in the 2020 UMD Data Challenge.

The initiative, hosted by the University of Maryland, challenged student teams to develop implementable, scalable data science projects that link to the EPA’s Three Pillars of Sustainability: environmental, social, and economic. Judges evaluated entries on criteria including innovation, teams’ final presentations, and results.

Contributing data for a greater cause

As a Gigabyte Sponsor for the 2020 event, Booz Allen mentored teams, scoped projects, provided technical support and guidance to student teams, and contributed three datasets to the initiative: 

• The “New Pollution” dataset identifies the time and location of different pollutants, for models that forecast the different concentrations of pollutants.
• The “Ocean Cleanup” dataset featured U.S. data representing zone, clean-up date, trash type, pounds of trash and more from TIDES, the world’s largest dataset representing ocean trash.
• The Recycling Diversion dataset is comprised of time series data for predicting recycling behavior.

Bringing people and information together for a healthier planet

These datasets contributed to several of the winning projects: Grand Prize, Outstanding Graduate Project, Best Global Sustainability Challenge, Outstanding UMD Project, and Outstanding Undergraduate/Graduate Project. Booz Allen Senior Consultant Alex Baker mentored one of the top award-winning teams: Yanzhi Shen and Zhenyang Wang, Grand Prize and Best UMD Project for their work with the New Pollution dataset.

Students were given a week to work on their projects, beginning with a day-long kickoff event and an in-person hack session with their assigned mentors on February 22, 2020. The event culminated with final presentations and judging on February 29, 2020.

“It was rewarding to see teams work diligently on their projects despite their busy schedules and lack of exposure to data science and the data science workflow,” Baker said. “The students were more excited than I expected and did not let their lack of experience slow them down.”

“Many were also excited to learn more about data science and software engineering at Booz Allen, so hopefully we will see them in the future as interns or full-time employees.”

Learn more about the UMD Data Challenge and data science at Booz Allen.