The 8 intersections controlled for the real-world experiments were divided into 3 groups based on the corridors on which they are located.
This dataset combines data from traffic sensors, accident data, weather data, and daylight information into a comprehensive tagged traffic accident dataset. This dataset contains …
Automated Incident Detection (AID) pipeline as described in the paper. After the dataset generation phase (which is described in detail in our Data in Briefpublication), we …
Real-world road (top) and VR digital twin versions under two different light conditions (middle and bottom).
Setup for dynamometer testing (left) and view from inside the vehicle showing simulated traffic linked to the driver, and a mobile app showing speed recommendations (right).
Cyber-physical control experiments conducted through the CTwin’s cyberinfrastructure: Connecting traffic light controller at intersections with signal timing optimization …
Model selection is performed using a data availability map showing which variables (cases, hospitalizations, deaths) each of the dozens of predictive models includes, and which …
Comparison of an animated trails visualization (used at microscopic scale) and our proposed animated strips-network for visualizing lane usage at mesoscopic scale for a section of …
The calendar view shows the incident distribution of a selected dataset over time. Each day’s incidents are displayed as a tile on the calendar. When the tile for a day is …
Chord charts for intersection-level visualization for a single day (Tuesday, April 2, 2019) in Chattanooga, Tennessee. The horizontally placed charts represent the intersections …