Objectives and Topics

One critical bottleneck to improving the safety performance of connected and automated vehicles (CAVs) is the difficulty of testing and evaluating safety performance. To date, there are no consensus or standard procedures on how to test and evaluate CAVs. During the past few years, although the problem of CAV testing has been investigated extensively by various CAV developers, government agencies, professional organizations, as well as academic institutions, the theories, methods, and best practices to support such testing and evaluation are still lacking. 

The 2nd workshop on safety testing and validation of CAVs focuses on sharing the state-of-the-art theories, methods, algorithms, metrics, datasets, simulations, and field implementation of safety performance testing and validation of CAVs, and identifies challenges as well as research needs, aiming to encourage cross-disciplinary cooperation and facilitate the development and deployment of CAVs.  The main topics of interest are listed below:

• Methodological frameworks for safety testing and validation

• Advanced data engines for large-scale, high-resolution, and rich-context data collection

• Naturalistic driving environment modeling with statistical realism

• Testing scenario generation for comprehensive behavior competency evaluation

• Testing environment generation for trip-level safety performance evaluation

• Adversarial testing and validation frameworks, adversarial attacks, and falsification of CAV sub-modules (e.g., perception, decision-making, communication, and control)

• Field implementation, simulation, and multi-fidelity (sim-to-real) frameworks for safety testing and validation

 Supported by