Diseño de un entorno de realidad virtual del juego La Escalera para niños con cociente intelectual limítrofe

Abstract

In the context of the Mathematics Education emphasis of the Interinstitutional Doctorate in Education (DIE), a virtual reality game titled “Parametric Ladder” has been developed for children with borderline intellectual functioning. The game was developed in Unity 6 for the Meta Quest 2 device, integrating educational mechanics and an intelligent feedback system based on artificial intelligence agents and graph analysis using optimized search techniques.

The system workflow operates as follows: when the player makes a move within the VR environment, it is sent to an API that records the move in a relational SQL database for basic persistence. Simultaneously, the system constructs and maintains a decision graph in JSON format (key–value structure) that represents learning trajectories in real time.

This graph is optimized to identify and prioritize the most common moves, significantly reducing processing time. A specialized decision-making module analyzes the path constructed by the player and contrasts it with a graph of precomputed possibilities, using search and optimization techniques (supported by DGX infrastructure for intensive computing) to determine the next best move and assign the most appropriate agent (provide feedback, advise, explain, demonstrate, or ask) for each situation.

The system’s response is transmitted back to the virtual reality environment in the form of audio or small guidance prompts, creating an adaptive and real-time learning experience. The hybrid architecture (SQL + JSON) enables both reliable storage and agile processing of learning trajectories, offering an immersive tool that combines virtual reality, AI, and graph theory to strengthen cognitive skills in populations with borderline intellectual functioning.