Data and the Bank of America Chicago Marathon


Data abound at the Bank of America Chicago Marathon. While runners track numbers related to their pace and distance, those in Forward Command, the heart of the marathon’s organization, keep an eye on logistical statistics related to the event’s operation: how many runners have passed each 5K mark? Where are participants on the course? What sort of medical resources are available at various aid stations? Thanks to a grant from the National Science Foundation, students in Northwestern University’s McCormick School of Engineering are harnessing the power of this data to help the race shape its management.

Since 2013, Professor Karen Smilowitz, professor of industrial engineering and management sciences at Northwestern, and Northwestern students have worked with the Bank of America Chicago Marathon to study medical logistics and preparedness in the marathon specifically, with the goal of using that information to apply it to other mass participation events.

“We’re looking at key data streams to give you situational awareness: where are runners? What is the load on medical resources?” Smilowitz says. “We’re able to predict where runners will be so if something were to happen and you had to stop the race, you know where the runners would be.”

The work done by Smilowitz and her students particularly focuses on medical tracking. Medical tents require volunteers to scan in and out, and along the course, tents use handheld tablets that automatically send information to Forward Command in real time. This helps the race and the Northwestern team know the location of resources and make adjustments if needed: for example, if one medical tent needs more staff, or which medical tents have the most traffic.

As the partnership has continued, however, the team from Northwestern has started to create race simulations that can assist those in operations with predicting how the race will progress.

“Students this year used historic marathon data going back to 2008 to simulate the marathon,” Smilowitz says. “If we know the weather on race day and know the demographics of the runners, know the corral assignments and pace of each corral, we can simulate the race. We use updates every 10 minutes as we get 5K counts and can adjust the simulation if people are going faster or slower than expected. This will be the first year of trying it out.”

Though race day data help those in Forward Command at the Bank of America Chicago Marathon the most, Smilowitz and her team have also worked with the Houston and Boston Marathons.

“Houston uses the same medical patient tracking system that originated in Chicago,” Smilowitz says. “The data comes in the same form, so you can analyze it. It’s great for us, because with the race in Chicago, you only get one set of data points per year.”

Smilowitz and her team have a couple years remaining on their grant and hope to continue this sort of data analysis moving forward.

“For me as a faculty member, it’s been a great partnership to get students out in the field and involved in such a great Chicago event,” Smilowitz says. “We’ve had a great opportunity here to work with Chicago, and it’s been really neat.”