Butler 209 will be hosting at least one thing this finals season.

A team of five sophomores from Columbia College and the School of Engineering and Applied Science is currently on course to endow the entrances to Butler 209 with bidirectional movement sensors to collect data for Seatz, a new Columbia Libraries seat-tracking app. An older app, CU Density, utilized WiFi data via campus routers to estimate the available space in various well-traveled areas, but has been plagued over the past years with spotty technical support, glitches, and UI issues.

Seatz, which instead utilizes data assembled from Raspberry Pi units and small processors installed at various doorways, will incorporate machine learning programs to analyze incoming data and estimate the capacity of each Library room. The largest challenge in such an endeavor, Project Leads Deniz Ulcay (SEAS ’19) and Lora Beltcheva (CC ’19) recounted in an interview this afternoon, involves the number of students who leave their materials in a room immediately after placing their belongings at an open space. While the bidirectional motion sensors will accurately record the number of students who have entered and exited the room, this statistic alone may not accurately imply how full the room is.

Ulcay’s team believes they “can statistically work around this,” given Butler Library’s Administration partnership with Ulcay in this project so far. In conjunction with the data analysis from the motion sensors, Butler Library’s VP for Digital Programs and Technology Services Robert Cartolano and Interim Librarian for Collections and Services Barbara Rockenbach have offered to collect real-time data on occupancy during Seatz’ beta testing period. “The libraries are trying to do a lot of stuff to destress students,” they explained, and “they’ve been quite supportive so far.”

Raspberry Pi unit with a banana for scale.

With Butler Library’s dedication and engagement, the project team’s goal to reach a “universal level” for tracking space occupancy throughout Columbia’s campus appears to be within grasp. “We are hoping to have a system applicable to any university or system of libraries…[and] leave a system that will be able to sustain itself,” Ulcay and Beltcheva stated. Specifically, the team hopes to place sensors at the doors to every room in the Library, keep track of how many enter and exit, and give a traffic report from a scale of “red to green.” They plan to use collected data to make analytical forecasts for the occupancy levels of the rooms, and then sort a list of Library rooms organized by inputted personal qualifications. This goal is perhaps in reach due to the extremely low cost of hardware installation and software support. A single Raspberry Pi unit can collect data from multiple doors simultaneously, while the processors themselves cost under $15 per unit. Butler Library can be entirely hooked up for “within three digits.”

The team has so far self-funded their project, and have maintained a degree of separation from any standard source of student project funding. While this “is not a project that should be connected to any of the student bodies,” Ulcay said, “at the same time…CCSC and ESC both have been very active and eager to find out and learn more.” ESC University Senator Izzet Kebudi, specifically, has been working with the Seatz team to arrange and organize appropriate meetings with Library staff. Once the team has finished the installation and beta testing for Butler 209, the “funding should come from the Libraries” for the remainder of the non-profit project.

You’ll be able to spot the first units sometime in the next two weeks in Butler 209.