Learning Latent Traits for Simulated Cooperative Driving Tasks
This work addresses the challenge of personalized AI interaction in cooperative driving for improved safety, though it is incremental as it builds on existing methods with a new simulation environment.
The paper tackled the problem of understanding individual human preferences and behaviors for effective human-AI teaming in risky driving scenarios by developing a framework that learns compact latent representations from simulated driver data, resulting in quantified discrimination of driver types and intervention policy effectiveness.
To construct effective teaming strategies between humans and AI systems in complex, risky situations requires an understanding of individual preferences and behaviors of humans. Previously this problem has been treated in case-specific or data-agnostic ways. In this paper, we build a framework capable of capturing a compact latent representation of the human in terms of their behavior and preferences based on data from a simulated population of drivers. Our framework leverages, to the extent available, knowledge of individual preferences and types from samples within the population to deploy interaction policies appropriate for specific drivers. We then build a lightweight simulation environment, HMIway-env, for modelling one form of distracted driving behavior, and use it to generate data for different driver types and train intervention policies. We finally use this environment to quantify both the ability to discriminate drivers and the effectiveness of intervention policies.