Project C.A.P.E. (Changkat AI Physical Education) is a pedagogical trial designed to optimize basketball skill acquisition across two pilot classes. By blending home-based video assessment with an in-class Station Rotation Model, the project delivers semi-automated, personalized feedback to students, allowing the teacher to execute high-impact Differentiated Instruction (DI) on the court.

Overview
Students are loaned basketballs home (1 each) to practise challenging but relevant ball control tasks at home. The task executed are filmed, uploaded on a AI/learning platform (SLS/Google Classroom) and mechanics will be applied for a semi-automated task analysis (much like SLS AI Feedback Assessments) for students to learn from.

When students report back to class, the lessons are organized by 3 levels of attainment achieved during the home-based practice tasks, and they go into 3-4 stations to learn at various pace in order to refine and improve on their game skills. The pedagogy in class is the station rotation model. We should observe increase in motivation and game skills excellence at the end of 6 weeks.

Problem Statement & Rationale
Traditional PE models often struggle to provide immediate, individualized biomechanical feedback to every student during a standard 70-minute lesson. Techniques that are intricate frequently go uncorrected, leading to the consolidation of poor movement patterns.

Theoretical Underpinning
According to Judith Rink’s Game Stages Theory, the bedrock of successful game skills acquisition is the control of objects, which she categorizes as Stage 1 of game play (Rink, 2020). Without minimal competency in object control, a student’s performance deteriorates rapidly when tactical complexities are introduced in later stages. Project C.A.P.E. uses home-based AI feedback to target and accelerate Stage 1 mastery, ensuring students are mechanically prepared to engage in the collaborative and tactical station rotations on the court.

Anchor Pedagogy
– Differentiated instruction,
– station rotation model with blended learning
– AI-assisted feedback semi-automation

Tools and platforms
– Singapore’s Student Learning Space (SLS)
– Gemini Pro (student edition)
– Google Classroom

Pedagogical Framework & Workflow
The trial operates on a cyclical, blended learning architecture distributed across a 6-week trial.