Using Explainable AI and Hierarchical Planning for Outreach with Robots

Making robot planning accessible builds AI literacy: The paper demonstrates that using explainable AI principles—such as visual interfaces and natural language explanations—can help non-experts (including K-12 students) understand how robots make decisions, addressing a key gap in public understanding of AI systems as they become more common in daily life.

Clear explanations of errors improve learning outcomes: JEDAI.Ed's approach of providing "helpful hints and natural language explanations for errors" significantly reduced the time needed for students to solve tasks compared to baseline methods, suggesting that transparency about why AI systems fail is crucial for effective human-AI interaction.

Personalized learning paths can democratize technical education: The platform's adaptive curriculum generation method customizes learning experiences for individual students, which the user study showed increased engagement—indicating that responsible AI education should account for different learning needs rather than using one-size-fits-all approaches.

Visual abstraction enables broader participation: By hiding complex technical details behind an intuitive visual interface while still allowing students to accomplish sophisticated tasks, the platform demonstrates that responsible AI development should prioritize making systems understandable to diverse audiences, not just experts.

User testing with non-experts validates real-world usability: The study specifically tested the platform with university students lacking computer science backgrounds, providing evidence that explainable AI tools actually work for their intended audience—an important validation often missing from AI research.