Introduction. Road traffic surveys based on visual field observations continue to be the primary method for collecting traffic flow data. However, in recent years, the adoption of automated video analysis software utilizing artificial intelligence technologies has gained significant traction. This approach is emerging as a viable alternative to traditional data collection techniques, offering enhanced efficiency and precision in measuring traffic parameters. By processing traffic video footage, it is possible to generate origin-destination matrices for specific vehicle types and analyze individual vehicle performance. This enables the collection of detailed traffic flow data, thereby improving the planning and management of road infrastructure.
Problem statement. The approach of utilizing traffic data video analysis with DataFromSky software by R.C.E. Systems can be effectively employed to calculate node impedance with intersection capacity analysis (ICA) within a transport macro model developed using PTV Visum software. However, to date, no studies have been conducted to assess the effectiveness of integrating these two technologies.
Purpose. Evaluation of the capabilities of software for automated video analysis of traffic data for modeling impedance in the nodes of a transport macro-model.
Materials and Methods. The study utilized traffic data video analysis powered by artificial intelligence, traffic indicator calculation based on the Highway Capacity Manual (HCM) methodology, and transport modeling of node impendence using ICA in PTV Visum.
Results. The case study of the controlled intersection at V. Matusevych Street and Metalurhiv Avenue in Kryvyi Rih revealed that the difference in traffic indicators obtained using ICA and DataFromSky (DFS) software was minimal. This suggests that the results from DFS analysis can be effectively used to calibrate the transport model developed in PTV Visum.
Conclusions. The indicators that can be assessed through traffic analysis results in DataFromSky include average total waiting time (delay), average level of service (LOS), load-to-capacity ratio, traffic volume (flow intensity), peak hour factor, maximum traffic volume in 15 minutes, and the proportion of right and left turns. While the vehicle queue indicators determined by ICA do not have a direct equivalent in DataFromSky, the Trajectories Movement Dynamics report generated by this software enables the establishment of relationships between speed, distance, and travel time for individual vehicles. Although ICA and DataFromSky employ different methodologies for analyzing traffic flow and intersection indicators, the results from video traffic data analysis can still be utilized for calibrating transport models and comparing model-based and real-world traffic indicators.
