A Methodology to Enhance Transparency for Trustworthy Artificial Intelligence for Cooperative, Connected, and Automated Mobility

@Article {
title = {A Methodology to Enhance Transparency for Trustworthy Artificial Intelligence for Cooperative, Connected, and Automated Mobility},
journal = {SAE International Journal of Connected and Automated Vehicles},
volume = {8},
keywds = {
CCAM; Data cards; Explainability; MLOps; Model cards; Transparency; Trustworthy AI
}
abstract = {

In this research, we propose a set of reporting documents to enhance transparency and trust in artificial intelligence (AI) systems for cooperative, connected, and automated mobility (CCAM) applications. By analyzing key documents on ethical guidelines and regulations in AI, such as the Assessment List for Trustworthy AI and the EU AI Act, we extracted considerations regarding transparency requirements. Recognizing the unique characteristics of each AI system and its application sector, we designed a model card tailored for CCAM applications. This was made considering the criteria for achieving trustworthy autonomous vehicles, exposed by the Joint Research Centre (JRC), and including information items that evidence the compliance of the AI system with these ethical aspects and that are also of interest to the different stakeholders. Additionally, we propose an MLOps Card to share information about the infrastructure and tools involved in creating and implementing the AI system.

}
doi = {10.4271/12-08-01-0010},
date = {2024-09-02},
}