In the previous column, the operational performance of an urban arterial nearby a driveway was demonstrated under various AV MPRs and traffic LOS (Mousavi et al., 2021). However, this column specifically focuses on the safety aspect of traffic on the urban arterial. The safety was analyzed through implementing Surrogate Safety Assessment Modes (SSAM) as well as driving volatility measures. The following sections cover SSAM results.
Surrogate Safety Assessment Model uses micro-simulation outputs to determine near-miss events using user-defined values for the predefined variables (Federal Highway Administration (FHWA), 2008; Pu & Joshi, 2008). Time-To-Collision (TTC) is one of the user-defined variables and is defined as the required time to collide if two vehicles continue on the same path with the same speed. Hence, in mixed traffic environments, various values should be used considering various perception reaction times of different vehicle types. Therefore, this research used two different TTC values to represent both conventional vehicles and AVs in mixed traffic environments. A value of 1.5 seconds was used to determine conventional vehicle conflicts (Federal Highway Administration (FHWA), 2008) and 1.0 second for AV conflicts (Morando et al., 2018).
Primarily, the value of 1.5 seconds was used as the TTC value to estimate the number of conflicts. As the next step, according to the type of vehicle that hit from the back, the TTC was adjusted. In fact, the SSAM has the capability of estimating the number of crossing, rear-end, and lane-changing conflicts. To accurately analyze the number of conflicts, only rear-end and lane-changing conflicts were considered in this research. In rear-end and lane-changing conflicts, the vehicle that hits from the back is known as the faulty vehicle. Therefore, if the vehicle that hit from the back was an AV, the TTC of 1.0 second was used, otherwise, the TTC value of 1.5 seconds was considered.
1.1. Rear-End Conflicts
Despite the operational analysis, the safety was evaluated for each scenario, regardless of the lane distribution, since conflicts are rare events. A previous study conducted by Mousavi et al. (Mousavi et al., 2020) assessed lane-based safety analysis. Table 1 provides the average number of rear-end conflicts for each LOS, AV MPR, and the vehicle at fault.
The results of the rear-end conflicts indicate that increasing the AV MPR results in a decrease in the number of conflicts. Table 2 proves that the average numbers of rear-end conflicts are significantly different for various AV MPRs. In fact, increasing the AV MPR significantly decreases the number of rear-end conflicts.
Figure 1, on the other hand, depicts how various traffic LOS influence the number of rear-end conflicts. Based on this figure, the average number of rear-end conflicts increases by increasing the traffic volume to LOS C due to increased traffic exposure. However, the LOS D results in a fewer number of rear-end conflicts compared to LOS B and LOS C. In fact, at LOS D, vehicles have less freedom to perform their turning maneuvers as it is harder to find an adequate number of gaps.
To enable a precise comparison between various traffic LOS, the number of conflicts was normalized by the traffic exposure. Figure 2 represents the normalized number of conflicts. The normalized number of the rear-end conflicts presents the same pattern as Figure 1.
In summary, increasing the AV MPR can significantly decrease rear-end conflicts while increasing the LOS increases the number of rear-end conflicts up to LOS C. At LOS D, fewer conflicts were observed.
1.1. Lane-Changing Conflicts
Similar to the rear-end conflicts, the TTC of 1.0 second was implemented if the second vehicle was an AV in lane-changing conflicts; otherwise, a TTC of 1.5 seconds was used. Figure 3 represents the total and normalized average number of lane-changing near-miss events. As indicated in Figure 3, by moving towards the LOS D, the number of lane-changing conflicts increases.
Figure 4 provides more details by representing the normalized number of lane-changing conflicts by traffic volume for each LOS and AV MPR. As depicted, for all the traffic LOS, the number of lane-changing conflicts increases by increasing the AV MPR to 10%. After that, any increase in the AV MPR almost steadily resulted in a fewer number of lane-changing conflicts.
On the other hand, increasing the traffic volume resulted in a decline in the number of LC conflicts until LOS C reached. However, at LOS D, the normalized number of LC conflicts escalated due to the merging of driveway vehicles to the arterial. These mandatory lane-changing conflicts shortened the spacing between the arterial vehicles and resulted in TTC values below the defined thresholds.
To conclude, regardless of the traffic LOS, increasing the AV MPR to any values over 10% enhances traffic safety by reducing the number of lane-changing conflicts.
Federal Highway Administration (FHWA). (2008). Surrogate Safety Assessment Model and Validation.
Morando, M. M., Tian, Q., Truong, L. T., & Vu, H. L. (2018). Studying the Safety Impact of Autonomous Vehicles Using Simulation-Based Surrogate Safety Measures. Journal of Advanced Transportation, 2018. https://doi.org/10.1155/2018/6135183
Mousavi, S. M., Dadashova, B., Lord, D., & Mousavi, S. R. (2020). Can Autonomous Vehicles Enhance Traffic Safety at Unsignalized Intersections? International Conference on Transportation and Development.
Mousavi, S. M., Osman, O. A., Lord, D., Dixon, K. K., & Dadashova, B. (2021). Investigating the Safety and Operational Benefits of Mixed Traffic Environments with Different Automated Vehicle Market Penetration Rates in the Proximity of a Driveway on an Urban Arterial. Accident Analysis & Prevention, 152. https://doi.org/https://doi.org/10.1016/j.aap.2021.105982
Pu, L., & Joshi, R. (2008). Surrogate Safety Assessment Model (SSAM): Software User Manual. May, 96. http://www.fhwa.dot.gov/publications/research/safety/08050/08050.pdf