Human-Centered Explainable Artificial Intelligence for Marine Autonomous Surface Vehicles

Explainable Artificial Intelligence (XAI) for Autonomous Surface Vehicles (ASVs) addresses developers’ needs model interpretation, understandability, and trust. As ASVs approach wide-scale deployment, these are expanded to include end user interactions in real-world contexts. Despite recent successes of technology-centered XAI enhancing the explainability AI techniques expert users, approaches do not necessarily carry over non-expert users. Passengers, other vessels, remote operators will have distinct from those users targeted a traditional approach. We formulate concept called ‘human-centered XAI’ address emerging interaction ASVs. To structure concept, we adopt model-based reasoning method formation consisting three processes: analogy, visualization, mental simulation, drawing examples ASV research at Norwegian University Science Technology (NTNU). The show how current activities point novel ways addressing underpin human-centered Findings representations (1) usability, (2) trust, (3) safety make up main processes XAI. contribution is help advance community’s efforts expand agenda interpretability, trust interactions.