Xiang “Jacob” Yan, Ph.D., an assistant professor in the Department of Civil and Coastal Engineering, is trying to make public transit more appealing to passengers and potential users.
Traditional approaches for bus stop amenities assessment often rely on manual surveys of each stop, which can be labor intensive. Instead, Yan and his research team developed a transit amenities assessment system (TAAS), which uses computers and artificial intelligence to quickly collect data for the public transit systems. The TAAS model integrates YOLOv8, a model great for identifying and pinpointing objects in real-time, along with a dynamic prediction algorithm and transfer learning to classify and segment Google Street View images.
“Transit agencies can use this system to create comprehensive, up-to-date inventories of bus stop amenities, enabling them to identify and prioritize necessary improvements. In fact, our ongoing research project aims to use the system to develop a nationwide transit stop amenities database,” Yan added.
Identifying the gaps in transit planning is crucial for a community’s population when they access public transportation, especially in major cities. The study, funded by the Center for Transit-Oriented Communities (CETOC), aims to build a structured system, picking up where previous data collections have left off. This approach focuses on amenities like shelters, seating, signage, lighting, and accessibility features at bus stops.
“These disparities highlight inequities in transit infrastructure planning, where certain areas – particularly those with lower ridership – may be underserved,” Yan said. “Cities can use this data to identify gaps in amenity provision at transit stops and prioritize improvements in underserved areas, particularly in neighborhoods with high transit dependency or vulnerable populations.”
In this initial study, Yan and the team selected five cities in Florida: Gainesville, Jacksonville, Miami, Orlando, and Tampa. Google Street View images from each bus stop were used to ensure the accuracy, efficiency, and cost-effectiveness of bus stop amenities assessment.
“The system can be deployed across different urban contexts with relatively low additional cost and effort, making it a practical solution for large-scale assessments of bus stop amenities,” Yan said. Beyond low-cost improvements, he said this effort is important because these “amenities make waiting safer and more comfortable, which can encourage people to use buses more.”
Yan is confident in his findings and knows that the automated assessment system will continue to identify other amenities such as trash bins, paved boarding areas, crosswalks, and pedestrian-level lighting to improve transit infrastructure projects.
After his research is completed, he believes that transit agency staff and transit advocates across cities can integrate the system to create comprehensive, up-to-date inventories of bus stop amenities to prioritize necessary improvements.
Yan and the research team plan to develop a nationwide transit stop amenities database through CETOC funding.
—
Written by Reba Liddy,
ESSIE Marketing and Communications Specialist