Pasar al contenido principal

EVALUACION NO DESTRUCTIVA Y MONITORIZACION DEL RADON UTILIZANDO TECNOLOGIAS NO DESTRUCTIVAS Y ROBOTICA EN EDIFICIOS DE PATRIMONIO HISTORICO (RADIANT-NDT)

PID2024-159276OB-C41

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal para la Investigación y el Desarrollo Experimental
Subprograma Subprograma Estatal de Generación de Conocimiento Científico-Técnico y Desarrollo Experimental
Convocatoria Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Año convocatoria 2024
Plan Estatal 2024-2027 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario AGENCIA ESTATAL CONSEJO SUPERIOR DE INVESTIGACIONES CIENTIFICAS (CSIC)

Resultados relacionados

Resultados totales (Incluyendo duplicados): 1
Encontrada(s) 1 página(s)

DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

RadonFAN code, RadonFAN - Intelligent Real-Time Radon Mitigation through IoT, Rule-Based Logic, and AI Forecasting

  • Abad Azcutia, Lidia
  • Ramonet Marques, Fernando
  • González Hernández, Margarita
  • Anaya Velayos, José Javier
  • Aparicio Secanellas, Sofía
The software is original. It relies on standard open-source libraries but is not derived from a specific existing codebase., [Relationship between files, if important] The CSV files in data/ are the primary input to the preprocessing scripts in sc/. These scripts generate training, validation, and test splits that are consumed by the model training routines. Trained model checkpoints are stored in models/ and can be loaded by evaluation or deployment scripts. Notebooks in notebooks/ provide illustrative analyses and reproduce figures and tables reported in the associated publication., [Running the tests (explain how to run the automated tests for this system. Explain what these tests test and why)] Testing scripts evaluate trained models on held-out test sets and compute classification metrics (precision, recall, F1-score, Brier score) as well as system-level indicators such as unsafety counter and energy consumption ratio. These tests ensure that models are suitable for safety-critical radon mitigation. To run test use the scripts above., [Deployment (additional notes about how to deploy this on a live system)] The trained DC model is lightweight and suitable for real-time inference on embedded platforms such as Raspberry Pi. Deployment scripts are not provided., RadonFAN is an open-source IoT and AI-based system for real-time indoor radon monitoring and mitigation. It integrates low-cost radon sensors, microcontrollers (Arduino and Raspberry Pi), cloud communication, and automated ventilation control. The software implements both AI-driven predictive models based on time-series classification (Direct Classification, DC, and Regression-to-Classification, R2C) to anticipate hazardous radon concentration exceedances and proactively activate ventilation systems. The code supports data acquisition, preprocessing, model training and evaluation scripts, and offers pretrained checkpoints. The current version corresponds to the models and experiments reported in the article “RadonFAN: Intelligent Real-Time Radon Mitigation through IoT, Rule-Based Logic, and AI Forecasting”., Grant RPID2024-159276OB-C41 funded by MICIU/AEI/10.13039/501100011033 and by ERDF/EU., File List: data/ – Available CSV files containing radon measurements and control variables (e.g., RG1.csv, RG2.csv). models/ – Checkpoints of the final trained deep learning models (DC and R2C). notebooks/ – Jupyter notebooks used for exploratory data analysis, visualization, and preliminary experiments. sc/ – Python scripts for data preprocessing, model training, validation, testing, and evaluation., Peer reviewed