A New Modeling Approach for Predicting Vehicle-Based Safety Threats

Abstract

Existing autonomous driving systems of intelligent vehicles such as advanced driver assistant systems (ADAS) assess and quantify the level of potential safety threats. However, they may not be able to plan the best response to unexpected dangerous situations and do not have the ability to cope with uncertainties since not all vehicles can always keep a safe gap from preceding vehicles and drive at a desired velocity. Previous research has not taken such uncertainties into account, it is, therefore, necessary to develop models which are not restricted by the predefined movement patterns of a vehicle. Existing systems are based on a model that estimates the threat level based only on one factor Time-To-Collision (TTC). This approach is limited since it cannot handle all scenarios and ignores all uncertainties. To overcome these limitations, this paper utilised deep learning to develop a range of models that rely on a group of factors to reliably estimate the threat level and predict conflicts under uncertainty using the concept of 'looming'. Comparative analyses were undertaken by incorporating new varying input factors to each model (e.g., surrogate safety measures, vehicle kinematics, macroscopic traffic data). Real-world experiments demonstrated that adding new factors increases the reliability and sensitivity of the models. Results also indicated that the models that consider looming provide low false alarm rate extending their applications for a wider spectrum of traffic scenarios. This is paramount for ADAS as uncertainties are inherent in the deployment of connected and autonomous vehicles in a mixed traffic stream.

Assessing temporary traffic management measures on a motorway: Lane closures vs narrow lanes for connected and autonomous vehicles in roadworks

“…Its equation is given as (Equation 4): $$$$\begin{equation}{\mathrm{TTC }} = \frac{{\mathrm{s}}}{{{v}_{\text{ego-vehicle}} - {v}_{{\mathrm{leading}}}}}\ \forall \ {v}_{\text{ego-vehicle}} > {v}_{{\mathrm{leading}}}\end{equation}$$$$where $${v}_{\text{ego-vehicle}}$$ is the speed of a CAV, $${v}_{{\mathrm{leading}}}$$ is the speed of the leading vehicle, and $${\mathrm{s}}$$ is the distance between preceding vehicle and the CAV. A TTC of 1.5 s is a widely adopted threshold in the literature as a safety surrogate measure [47–51] and is used as a benchmark for comparison purposes. Similarly, to evaluate the safety of CAVs in this context, (TTC) metric was also adopted with a threshold of 1.5 s. This threshold is widely used in safety assessments of automated driving systems [52, 53].…”

“…where v ego-vehicle is the speed of a CAV, v leading is the speed of the leading vehicle, and s is the distance between preceding vehicle and the CAV. A TTC of 1.5 s is a widely adopted threshold in the literature as a safety surrogate measure [47][48][49][50][51] and is used as a benchmark for comparison purposes. Similarly, to evaluate the safety of CAVs in this context, (TTC) metric was also adopted with a threshold of 1.5 s. This threshold is widely used in safety assessments of automated driving systems [52,53].…”

Real-time Road Traffic Data Monitoring using Novel Leaky-Wave Antennas in a Connected and Autonomous Vehicles Environment

“…Additionally, the data collected is gathered at a minute-level granularity. However, for future applications the disaggregated data goal is to minimise the prediction horizon, aiming for intervals of less than 1 second for CAVs, intelligent transportation systems, and vehicular communication (19,20).…”

Safety in Traffic Management Systems: A Comprehensive Survey

“…The paper identified opportunities for integrating computer vision with other technologies and methods for safety enhancement. Formosa et al [80], in their research, directed their attention towards advanced driver assistant systems (ADAS) and highlighted the uncertainties associated with the deployment of connected and autonomous vehicles into heterogeneous traffic environments. They emphasized the limitations of previous studies that predominantly relied on predefined movement patterns and a single factor (time to collision) to estimate the threat levels.…”