Using Artificial Intelligence and Virtual Simulators as a Pedagogical Strategy in Teaching Robotics

Abstract

Poor academic performance in technology-oriented courses is a recurring problem in Latin American higher education. In Robotics I, offered by the Telecommunications Engineering Program at Universidad San Francisco Xavier, this problem manifests as a 69.2% failure rate across two consecutive academic terms and an average grade of 46.08 out of 100, with no formal integration of artificial intelligence in teaching strategies. The study aimed to analyze the use of artificial intelligence and virtual simulators as a pedagogical strategy, and to design and validate a proposal aimed at strengthening the teaching–learning process in this course. A mixed quantitative-qualitative approach was adopted, with a non-experimental cross-sectional design framed within the socio-critical paradigm. Five instruments were applied (two questionnaires, two interview guides, and one document analysis guide) to 20 students and 4 instructors. Source triangulation revealed a 2.45-point gap in AI use between active users (mean 4.29) and non-users (mean 1.84). Based on that diagnosis, a strategy was designed integrating six tools (Wokwi, TinkerCAD, Cisco Packet Tracer, RoboDK, Webots, and Codeium) within a four-phase model grounded in Kolb's experiential learning theory. The proposal was validated through the Delphi method with a panel of 10 experts; all items received ratings of "Quite adequate" or "Very adequate", with no "Inadequate" judgments, confirming its relevance and institutional viability.