Project Details
Description
The present proposal aims to develop a comprehensive assessment of the radiation protection program of the Centro de Investigación y Extensión en Ingeniería de los Materiales (Research and Extension Center for Materials Engineering, CIEMTEC) of the Instituto Tecnológico de Costa Rica (Costa Rica Institute of Technology), with the purpose of evaluating the degree of compliance of the facilities that use ionizing radiation with the technical and administrative requirements established in Executive Decree No. 44653-S. Laboratories that use radioactive sources or ionizing radiation generators in their activities require both documentary and operational updates, as their procedures were developed under regulations that predate the current regulatory framework.
The methodology integrates regulatory, technical, and operational evaluation through the complementary use of Artificial Intelligence tools applied to natural language processing (NLP) and data analysis. For the documentary analysis, AI-NLP models are employed to perform semi-automated comparisons between regulatory requirements and the documentation of the Research Center, allowing the identification of gaps, deviations, and opportunities for technical standardization. At the operational level, the diagnostic assessment includes the review of aspects related to infrastructure, shielding, signage, dosimetry, maintenance, training, and emergency response, supported by AI-assisted data analysis techniques, Python-generated scripts, statistical analysis, and preliminary modeling of bunker conditions through comparison of specialized software.
The results will be consolidated into a diagnostic report that synthesizes findings, compliance gaps, and recommendations aimed at optimizing radiological management, strengthening the safety culture, and guiding improvement actions. The use of AI increases the efficiency of the analysis without replacing technical review, providing methodological innovation and enhancing institutional capacity for informed decision-making in matters of radiological safety.
The methodology integrates regulatory, technical, and operational evaluation through the complementary use of Artificial Intelligence tools applied to natural language processing (NLP) and data analysis. For the documentary analysis, AI-NLP models are employed to perform semi-automated comparisons between regulatory requirements and the documentation of the Research Center, allowing the identification of gaps, deviations, and opportunities for technical standardization. At the operational level, the diagnostic assessment includes the review of aspects related to infrastructure, shielding, signage, dosimetry, maintenance, training, and emergency response, supported by AI-assisted data analysis techniques, Python-generated scripts, statistical analysis, and preliminary modeling of bunker conditions through comparison of specialized software.
The results will be consolidated into a diagnostic report that synthesizes findings, compliance gaps, and recommendations aimed at optimizing radiological management, strengthening the safety culture, and guiding improvement actions. The use of AI increases the efficiency of the analysis without replacing technical review, providing methodological innovation and enhancing institutional capacity for informed decision-making in matters of radiological safety.
General Objective
Realizar un diagnóstico de situación del programa de protección radiológica del CIEMTEC mediante la evaluación técnica, normativa y operativa de sus instalaciones que utilizan radiaciones ionizantes, apoyándose en herramientas de Inteligencia Artificial (IA) en el análisis documental y de datos, para la determinación del nivel de cumplimiento con la legislación vigente y la proposición de acciones de mejora alineadas con los principios de seguridad y optimización radiológica
Research Lines
Física Aplicada
| Status | Active |
|---|---|
| Effective start/end date | 1/01/26 → 31/12/26 |
Fingerprint
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.