A User-Centred Framework for Explainable Artificial Intelligence in Human-Robot Interaction

XAI is fundamentally a communication problem, not just a technical one. The paper emphasizes that explainability requires understanding how humans actually process and comprehend information, drawing on cognitive and social sciences rather than relying solely on computer science approaches.

One-size-fits-all explanations don't work for diverse users. The framework prioritizes tailoring XAI solutions specifically for non-expert users, recognizing that general audiences need different types of explanations than AI specialists or researchers.

Real-world AI deployment demands user-centered design. As AI systems move from laboratories into practical applications (like robotics), the paper argues that explainability solutions must be designed with actual end-users in mind from the start, not added as an afterthought.

Interactive explanation is more effective than static transparency. The framework focuses on the "social-interactive aspect" of XAI, suggesting that two-way communication between humans and AI systems produces better understanding than simply making algorithms more transparent.

Responsible AI requires interdisciplinary collaboration. The paper demonstrates that creating trustworthy, understandable AI systems requires insights from cognitive science, social science, and human-computer interaction—not just computer scientists working alone.