This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Educational robotics, powered by artificial intelligence (AI), represents a significant transformation in teaching and learning environments, especially in STEM areas. This study aims to analyze, through a systematic review, the advances, applications, and challenges of educational robotics integrated with AI, identifying pedagogical trends, research gaps, and emerging approaches between 2020 and 2025. The PRISMA protocol was applied, complemented by elements of Grounded Theory, to ensure methodological rigor and interpretive depth. The search was conducted in academic databases such as Scopus, ScienceDirect, ERIC, SciELO, Redalyc, and KM&EL, using standardized descriptors and Boolean operators. After applying inclusion and exclusion criteria, 14 peer-reviewed scientific articles were selected. The results show that educational robotics with AI improves student motivation, active learning, and retention of scientific concepts by up to 30%. Robots have proven effective in special education contexts, promoting social-emotional skills. However, 43% of the studies lack solid pedagogical foundations, and 40% of institutions face technological barriers that limit equity in access. It is concluded that educational robotics with AI has high transformative potential, provided that it is articulated with active methodologies, continuous teacher training, and inclusive public policies. This review contributes to a critical understanding of the field, promoting the design of more personalized, ethical, and sustainable learning experiences.
