"IA para cerrar la brecha digital en Perú: Una propuesta de mapeo de zonas no conectadas"

Abstract

"The present article proposes the development of an Artificial Intelligence (AI)based strategy to map the unconnected areas of Peru using prioritization criteria, in order to improve the planning of public and private investment in telecommunications. This strategy seeks to overcome the limitations derived from the use of scattered data and to more accurately identify the localities that should be prioritized to efficiently guide resources and public policy decisions. [Methodology/approach/design] The strategy is composed of three main components. First, the identification of unconnected areas through a system that integrates computer vision and machine learning techniques, considering prioritization criteria. Based on the calculation of the Priority Index (PRI), 11,218 localities were detected. Second, the clustering of these localities through a spatial analysis process using the DBSCAN algorithm, which allows the detection of clusters of underserved areas that share geographical proximity and similar characteristics. The system identified 160 clusters distributed mainly in regions of the Amazon (Loreto, Amazonas, San Martín) and the Andes (Puno, Junín). [Findings] Finally, the prioritization of public or private interventions based on these results makes it possible to identify —by selecting the 5% of localities with the highest priority— 2,824 CCPP, a potential benefit for approximately 389,445 people. [Practical implications] This provides a tool that strengthens the strategic planning of the Ministry of Transport and Communications - MTC, the Supervisory Body for Private Investment in Telecommunications - OSIPTEL, and the National Telecommunications Program - Pronatel, by enabling the definition of coverage obligations with greater precision, scope, and positive impact on citizens through instruments such as the coverage canon, the renewal of concessions, or radio spectrum contests. It is considered essential that the entire strategy be developed under principles of transparency, ethics, accountability, and compliance with the Peruvian regulatory framework, in particular Law No. 31814 and its regulations, ensuring that the use of AI responds to the public interest. [Originality/value] Along these lines, it is expected that the strategy, by being systematic, reproducible, and innovative, by integrating geospatial analysis and machine learning to identify, cluster, and prioritize unconnected areas with greater precision, constitutes a disruptive approach compared to traditional methods. Thereby, it would contribute to more efficient and equitable decisions to close the digital divide in Peru. © 2026 Universidade de Brasilia. All rights reserved."