RoboBlockly Studio: Conversational Block Programming With Embodied Robot Feedback for Computational Thinking

Synthesis: Li, Du, Sun, and colleagues (2026) design and evaluate RoboBlockly Studio, an integrated interactive system that combines block-based programming, a conversational AI teaching agent, and embodied robot execution to support computational thinking. Recognizing that learners and teachers face challenges connecting abstract program logic to meaningful outcomes, the system creates a tight iterative loop of authoring, running, observing, and revising. Informed by interviews with five programming teachers, the system was designed to support four goals: preserving learner agency, making program behavior transparent, grounding programming in embodied classroom-aligned tasks, and Scaffolding reflection through AI dialogue. It was deployed with 32 high school students.

Key Findings

• Computational thinking (CT) is increasingly promoted as a core literacy, yet learners and teachers face challenges connecting abstract program logic to meaningful outcomes.
• RoboBlockly Studio combines block-based programming, a conversational AI teaching agent, and embodied robot execution, creating a tight iterative loop of authoring, running, observing, and revising.
• Informed by interviews with five programming teachers, the system supports four goals: (1) preserving learner agency, (2) making program behavior transparent and interpretable, (3) grounding programming in embodied, classroom-aligned tasks, and (4) scaffolding reflection through pedagogically grounded AI dialogue.
• The system was deployed with 32 high school students, with observation of how robot and AI dialogue affected learning.

Study Design & Method

This is a design-based development and deployment study. The researchers designed RoboBlockly Studio through iterative consultation (interviews with five programming teachers informed the four design goals), combining block-based programming with a conversational AI teaching agent and embodied robot execution. The system was deployed with 32 high school students, and the researchers observed how robot execution and AI dialogue supported computational thinking, including learner agency, program-behavior transparency, and reflection.

What this means for practice

• Teachers. Make program behavior visible during debugging: students used robot collisions and unintended detours as perceptual cues to localize failure, then corrected the specific block — as when P10 traced execution step by step and found a misplaced turn block rather than the end of the program.
• Teachers. Pair an embodied run with Execution Trace and Step-by-Step supports when the robot is unavailable, because learners then relied on stepwise inspection plus AI feedback to find where their reasoning diverged from expected behavior.
• Instructional designers. Route student errors into an AI Hint or Error Feedback loop that points to the program segment to revise, so learners treat mistakes as actionable signals for refining both code and strategy instead of receiving the corrected program.
• Instructional designers. Harden LLM-based checking before classroom use: the checker occasionally flagged functionally correct but unconventional block sequences as incorrect, so verify flagged solutions before they discourage valid CT strategies.

Limitations

The deployment involved 32 high school students, and detailed learning-outcome data are not fully reported in the abstract; the focus is on the design and observed use of the system. The system's effectiveness relative to other computational-thinking approaches requires comparative evaluation. Findings are specific to the high-school context and the particular robot/AI configuration.

Citation

Li, L., Du, C., Sun, J., et al. (2026). RoboBlockly Studio: Conversational block programming with embodied robot feedback for computational thinking.