Research Objective AIMATS allows for long-term traffic monitoring. Use of AIMATS leads to the collection of safety-relevant scenarios which can in turn be simulated to perform safety effectiveness assessment of advanced driver assistance systems (ADAS) and automated driving systems (ADS). Since its development in 2016, it has been used in numerous research projects. Until now processed data is exclusively based on German traffic accident data. This paper presents the results of using AIMATS in France, and thus allows comparison of speed data of specific traffic scenarios to those of Germany. Methodology Previous German AIMATS studies confirmed the hypothesis that critical driving situations occur more frequently at traffic accident blackspots. A research collaboration with CEESAR allowed for the expansion to France. The identification of French accident blackspots was executed by assessing cases from the French national road traffic accident database (BAAC). Personal injury accidents that occurred between 2011 and 2018 involving at least one passenger car were selected. Two AIMATS datasets, each involving multiple intersections, were compared: German data from 2017 and French data from 2021. For both datasets a classification of intersection types (T and X) and intersection maneuvers (by turning direction) was performed. Intersection traffic was grouped by both characteristics to determine vehicle speeds. The post-processing involved the comparison of speed data for different intersection types and maneuvers. Results For the French dataset, 4 blackspots were identified, corresponding to accidents that happened at intersections and including at least one accident with serious injuries (hospitalization for more than 24 hours). The German dataset was larger with 16 blackspots. The recorded data at the blackspots for both countries showed that the vehicle speed data correlated well for most combinations of intersection type and maneuver. Maneuvers with differences were identified and evaluated. In general, driving speeds were at or below 30 km/h at the apex of the curve for all intersection maneuvers. Limitations The selection of accident blackspots through AIMATS is generally based on collision characteristics that describe an accident such as the type of the accident, the type of involved participants, the intersection type, and its location. These attributes are available for the German data. Different characteristics and attributes in the French database due to different coding rules require an adapted selection of blackspots and thus measuring locations. In addition, the amount of video data collected in Germany was about ten times higher than in France. New: The paper provides novel results of recent AIMATS enhancements. This includes a comparison of the required input data and the results of the analysis between two countries. This is the first time that AIMATS was used in another country, indicating that this method is applicable in even more countries and regions of the world. Conclusion Further developments of AIMATS enabled its transfer to France. The analyses of German and French datasets resulted in maneuver-specific ranges of vehicle speeds and generally showed similar results between the two countries. This implies that the AIMATS approach can be applied to measurement sites in countries other than Germany.
Mr. Dominik Schreiber, Research assistant, Fraunhofer Institute for Transportation and Infrastructure Systems IVI