An X-in-the-Loop (XIL) Testing Framework for Validation of Connected and Autonomous Vehicles

Validation methods for growing connected and autonomous vehicle (CAV) technologies hold the key to their real-world implementation. Setting these up is challenging owing to the number of components and safety concerns involved in CAV testing. This paper presents a modular XIL framework for the validation of CAVs. The framework allows the interfacing of an in-house instrumented drive-by-wire vehicle with roads, traffic microsimulations (virtual vehicles, traffic lights, etc.), autonomy software, computing resources, and human drivers through the use of a mixed-reality headset. Three case studies in recent and ongoing work are shown, demonstrating the suitability and versatility of the XIL framework for verifying the safety and efficiency of automated vehicle control strategies via vehicle-in-the-loop, as well as ensuring realistic human driver reactions inside of emergent mixed-autonomy traffic environments via humans-in-the-loop.