Investigation of Key Safety Measures for Pre and Post-Deployment of Connected and Automated Vehicles

University

University of Kansas

Principal Investigator

Alexandra Kondyli (akondyli@ku.edu)

Total Project Cost

$346,243

Funding Type

2023 USDOT

Start Date

6/1/2023

End Date

7/31/2024

Agency ID or Contract Number

69A3552348307

Abstract

"This project will focus on understanding how safety-related metrics will likely change after the deployment of connected and automated vehicles (CAVs). As part of this project, first we will investigate the capabilities of the CAVs that are currently developed to date. Next, we will identify which safety metrics and associated performance measures, will be impacted by the deployment of new technologies, such as number of voluntary or mandatory take overs, crash frequency, etc. Next, we will investigate the operational domains and scenarios that these measures will be impacted and determine the magnitude of expected impact. We will also estimate the expected thresholds of the safety performance metrics. For example, the time to collision (TTC) threshold may be significantly lower for CAV/human-driven vehicles than for pairs of human-driven vehicles. The findings of this study will be important for state DOT or local agencies to evaluate the safety performance of their highway systems."

Description

"SSM used for both CAV and HDV. The literature will also summarize the CAV technologies that were assumed in past research, as well as operational domains. Finally, the literature review will also include a detailed description of the types and the characteristics of the crash prediction models used in SSM and SSM-based research. Task 2: Data Collection and SSM Analysis: During this task, the team will obtain trajectory data from online available datasets that include limited CAV data. An example of such dataset is the Waymo data (https://waymo.com/open/), which contains car following pair trajectories between a fully automated vehicle (AV) and HDV or between HDV. This is a good candidate dataset as it involves multiple geometric and environmental conditions for 1,000 trajectory pairs at 0.1s intervals. The research team will also try to obtain data that include interconnected vehicles. One example is testbeds that the Federal Highway Administration (FHWA) is currently deploying. Several different SSM will be calculated based on the available data. Task 3: Simulation of Specific Conditions: During this task, the research team will use simulation to generate additional vehicle trajectories for all three pairs, for specific geometric and environmental domains (e.g., freeways under good weather and sunny conditions). The simulation models will be calibrated using the field data. The Surrogate Safety Assessment Model (SSAM) will be used to analyze the trajectories and compute SSM. A comparison between the SSM that include CAV or only HDV will be conducted to evaluate the differences and capture the impact on automation on conflict research. Task 4: Final Report: The research team will compose the Final Report that describes the efforts undertaken during the final stage of the project."

Objective

"The expected results of this research project are a set of new or modified surrogate safety measures and their respective threshold values, specifically for connected and automated vehicles."

Impacts/Benefits

"Undergraduate and graduate students at the University of Kansas will be trained on highway safety topics. Students will be able to better understand safety analysis procedures as well as the implications of CAVs on safety. The findings of this research can also be used to inform policies related to the adoption of CAV technology and provide realistic expectations on the safety benefits stemming from CAVs."