LyndHub
Jadwal Sholat

Memuat jadwal sholat…

← All journalsSearch index

Ilmu Komputer & AI editorial

Open AccessOA2026

Teaching Reinforcement Learning and Humanoid Robotics to High-School Students: An Expert-Validated Curriculum Design on a Low-Cost Open Platform

An integrated, expert-reviewed framework that adapts a full robotics research workflow—assembly, simulation, RL policy learning, system identification, and physical deployment—into a coherent precollege course built around a shared open-source humanoid robot.
Yuanzhe Dong; Jie Cao; Shuman Wang· 2026· DOI 10.48550/arXiv.2609.25674

The core problem

Lower-cost open-source robots and reinforcement learning (RL) simulation tools have created new opportunities for precollege students to engage with contemporary robotics. However, translating a complete research workflow—spanning mechanical assembly, electrical setup, simulation, policy learning, system identification, and physical deployment—into a coherent course for novice learners remains challenging. The authors address this gap by presenting an integrated robotics course framework that organizes these activities around a shared robotic artifact. The framework is illustrated through a high school curriculum in which pairs of students assemble an open-source humanoid robot, train a walking policy in simulation, and deploy it on the physical platform. The work is motivated by the need for structured approaches that adapt authentic robotics research workflows into interdisciplinary precollege courses, while managing the cognitive and logistical constraints of novice learners.

Innovation

Expert feedback highlighted three central design tensions that shaped the final framework:

1. **Authenticity versus cognitive load**: Experts noted that while authentic research tasks are motivating, they can overwhelm novice learners if not carefully scaffolded. The framework addresses this by breaking the workflow into manageable stages and providing layered checkpoints.
2. **System integration versus timely visible progress**: Integrating multiple subsystems (mechanical, electrical, software) is essential for a holistic understanding, but it can delay visible progress. The framework mitigates this by sequencing activities to produce early, tangible milestones (e.g., assembling a subcomponent or achieving a simulated walking gait) before full integration.
3. **Team construction versus individual accountability**: Collaborative pair work is valuable for learning and motivation, but it can obscure individual contributions. The framework incorporates individual accountability structures, such as individual checkpoints and reflections, to ensure each student masters the material.

The final framework integrates these resolutions into a coherent course design. The curriculum was illu

Lower-cost open-source robots and reinforcement learning (RL) simulation tools have created new opportunities for precollege students to engage with contemporary robotics. However, translating a complete research workflow—spanning mechanical assembly, electrical setup, simulation, policy learning, system identification, and physical deployment—into a coherent course for novice learners remains challenging. The authors address this gap by presenting an integrated robotics course framework that organizes these activities around a shared robotic artifact. The framework is illustrated through a high school curriculum in which pairs of students assemble an open-source humanoid robot, train a walking policy in simulation, and deploy it on the physical platform. The work is motivated by the need for structured approaches that adapt authentic robotics research workflows into interdisciplinary precollege courses, while managing the cognitive and logistical constraints of novice learners.
The framework was developed through an iterative design process that included formative review by five experts in robotics research, engineering, secondary STEM education, and curriculum design. The design process combined several structural elements:

Why it matters

The paper contributes a design framework that bridges the gap between advanced robotics research and precollege education. By organizing activities around a shared robotic artifact, the framework creates a unifying thread that connects otherwise disparate disciplinary tracks. The three design tensions identified by experts are not unique to robotics education; they reflect broader challenges in STEM curriculum design, particularly when attempting to introduce authentic research practices to novice learners.

The emphasis on progressive integration of simulation and hardware is particularly noteworthy. Simulation provides a safe, low-cost environment for experimentation, while physical deployment introduces real-world complexities such as sensor noise, actuator limitations, and unmodeled dynamics. The use of system identification to bridge this gap is a sophisticated practice that is rarely introduced at the precollege level, yet the framework makes it accessible through scaffolding and checkpoints.

The framework's focus on balancing collaboration and individual accountability addresses a common pitfall in project-based learning. By incorporating individual checkpoints, the design ensures that all students develop core competencies even when working in teams. This is especially important in interdisciplinary contexts where students may have varying prior knowledge.

Limitations include the lack of classroom implementation data; the paper explicitly states that future studies are needed to examine implementation and student learning. Additionally, the framework was validated by a small panel of five experts, which may limit generalizability. Nonetheless, the work provides a valuable template for educators and researchers seeking to adapt robotics research workflows into precollege courses. The framework's modularity and reliance on low-cost open-source platforms suggest potential for broader adoption, though further research should investigate scalability, cost, and long-term student outcomes.

Who should read this

CS practitioners and researchers

Opening member content…