Project: #356 Bus on the Edge: Applications Progress Report - Reporting Period Ending: Sept. 30, 2022 Principal Investigator: Christoph Mertz Status: Completed Start Date: July 1, 2021 End Date: July 31, 2022 Research Type: Applied Grant Type: Research Grant Program: FAST Act - Mobility National (2016 - 2022) Grant Cycle: 2021 Mobility UTC Progress Report (Last Updated: Sept. 28, 2022, 7:31 p.m.) % Project Completed to Date: 100 % Grant Award Expended: 100 % Match Expended & Document: 0 USDOT Requirements Accomplishments The main goal of this project is to create application on the BusEdge system that are useful for the transit agencies and other stakeholders. The detection and analysis of when passengers enter and exit the bus is completed. We have found several instances in our data set where the bus stops are not marked in the official bus map. We had a meeting with the transit agency to get feedback on these results. For the second application we have investigated the best way to detect trash cans and their status. We noticed that there are two kinds of "full trash can". One is where the trash reached the top of the trash can and the second is where there is a full trash bag next to the trash can. Both kinds need attention. We have developed a 2-step detector which first detects the trash can and then determines the status of the trash can. The detector was tested in live operations on our BusEdge system. The work resulted in a RISS student poster and paper. The next application we developed is determining if the sidewalk near a bus stop is covered with snow. That requires to detect the location of the sidewalk during good weather and matching it with detection of snow during bad weather. The matching was done with COLMAP. The snow detection work resulted in another RISS student poster and paper. Another application we worked on was detecting changes to update high-definition maps. The particular map feature we observed was pedestrian crossings. This work resulted in a master's thesis. The final application we developed was methods for extracting the spatial attribute and the temporal attribute of construction zones from BusEdge images. This work also resulted in a master's thesis. Impacts If successful, our BusEdge system will help transit agencies to better manage their operations and assets. Results are discussed with transit agencies and bus equipment manufactures. Other We developed techniques to detect and analyze construction zones, HD map changes, full and empty trash cans, and snow covered sidewalks. We have collected a large set of videos from 5 cameras around a bus during regular transit operations covering more than one full year. We have open-sourced some of our software and some of our labeled data. Outcomes New Partners Hawaii transit agency TheBus (http://www.thebus.org/) has agreed to partner with us on the BusEdge project. Issues Nothing to report