Triangulating the future: Developing scenarios of cyclist-automated vehicle interactions from literature, expert perspectives, and survey data

Automated vehicles pose a unique challenge to the safety of vulnerable road users. Research on cyclist-automated vehicle interaction has received relatively little attention compared to pedestrian safety. This exploratory study aims to bridge this gap by identifying cyclist-automated vehicle scenarios and providing recommendations for future...

Effects of action space discretization and DQN extensions on algorithm robustness and efficiency: How do the discretization of the action space and various extensions to the well-known DQN algorithm influence training and the robustness of final policies under various testing conditions?

Reinforcement Learning (RL) has gained atten-tion as a way of creating autonomous agents for self-driving cars. This paper explores the adap- tation of the Deep Q Network (DQN), a popular deep RL algorithm, in the Carla traffic simulator for autonomous driving. It investigates the influ- ence of action space discretization and DQN ex-<br/...

Automated Repair of Feature Interaction Failures in Automated Driving Systems

In the past years, several automated repair strategies have been proposed to fix bugs in individual software programs without any human intervention. There has been, however, little work on how automated repair techniques can resolve failures that arise at the system-level and are caused by undesired interactions among different system...

Quantifying Take-Over Quality: A Novel Approach Utilizing a Scenario-Specific Optimized Reference Trajectory

Due to recent advancements in automated driving systems, drivers are able to withdraw themselves from the control loop of the vehicle in certain driving situations. However, system limits occur relatively frequently, which results in the driver having to take back control. It is argued most currently used take-over quality quantification methods...

Multi-Agent Deep Reinforcement Learning for Automated Highway Driving

Recent advances in Deep Reinforcement Learning have sparked new interest in many different research topics, including Automated Highway Driving where agents model autonomous vehicles. The main advantage of Deep Reinforcement Learning is that the training algorithm is adaptable to its environment. In highway driving, researchers often simplify...

A multi-level model on automated vehicle acceptance (MAVA): a review-based study

Automated vehicle acceptance (AVA) is a necessary condition for the realisation of higher-level objectives such as improvements in road safety, reductions in traffic congestion and environmental pollution. On the basis of a systematic literature review of 124 empirical studies, the present study proposes MAVA, a multi-level model to predict...

Human behaviour with automated driving systems: A quantitative framework for meaningful human control

Automated driving systems (ADS) with partial automation are currently available for the consumer. They are potentially beneficial to traffic flow, fuel consumption, and safety, but human behaviour whilst driving with ADS is poorly understood. Human behaviour is currently expected to lead to dangerous circumstances as ADS could place human...

Meaningful Human Control over Automated Driving Systems

Human Factors issues with automated driving systems (ADS) are becoming more apparent with the increasing prevalence of automated vehicles on the public roads. As automated driving demands increased performance of supervisory skills of the driver, rather than vehicle handling skills, a mismatch occurs between the demand and supply of the drivers’...

Meaningful Human Control over Automated Driving Systems

The Social Acceptance of Automated Driving Systems: Safety Aspects: A contribution to responsible innovation by using a referendum format, discrete choice model experiment to measure the social acceptance of ADS by Dutch citizens with corresponding heterogeneity

Automated driving systems (ADS) can improve traffic safety, improve accessibility and reduce environmental impact (Shladover, 2016). On the contrary, on May 7th 2016, a fatal accident with a Tesla on autopilot in U.S. Florida was a harsh reminder that the technology is still in its testing phase (Greenemeier, 2016). In complex technical systems...