The impact of occupancy dynamics on indoor CO2 forecasting: a cross-scenario evaluation

Indoor CO2 forecasting supports proactive ventilation control that balances air quality with energy efficiency. While Machine Learning (ML) models have shown strong performance in controlled settings such as schools, their generalization across indoor spaces with diverse occupancy dynamics remains p...

Descripción completa

Detalles Bibliográficos
Autores: Garcia-Pinilla, Peio, Jurío Munárriz, Aránzazu, Figols, María, Paternain Dallo, Daniel
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2026
País:España
Institución:Universidad San Jorge (USJ)
Repositorio:Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
OAI Identifier:oai:dnet:academicae__::0f662a392a2ef08611d8cc4bea5342c7
Acceso en línea:https://hdl.handle.net/2454/57128
Access Level:acceso abierto
Palabra clave:Indoor air quality
Forecasting
Machine learning
Foundation models
Indoor CO2
IoT
id ES_acd9fdc8b539286ee771c0c38fddff53
oai_identifier_str oai:dnet:academicae__::0f662a392a2ef08611d8cc4bea5342c7
network_acronym_str ES
network_name_str España
repository_id_str
spelling The impact of occupancy dynamics on indoor CO2 forecasting: a cross-scenario evaluationGarcia-Pinilla, PeioJurío Munárriz, AránzazuFigols, MaríaPaternain Dallo, DanielIndoor air qualityForecastingMachine learningFoundation modelsIndoor CO2IoTIndoor CO2 forecasting supports proactive ventilation control that balances air quality with energy efficiency. While Machine Learning (ML) models have shown strong performance in controlled settings such as schools, their generalization across indoor spaces with diverse occupancy dynamics remains poorly characterized. We present a systematic benchmark of 11 forecasting models spanning simple baselines, statistical methods, classical ML, deep learning, ensembles, and foundation models using 18 weeks of IoT sensor data spanning six real-world use cases: conference rooms, dining halls, hospitals, food markets, offices and student residences. Performance depends strongly on the prediction horizon and on the regularity of occupancy-driven CO2 patterns. Simple baselines tend to perform best at short horizons (10 min ahead), while ensembles and fine-tuned foundation models provide more robust accuracy at longer horizons (4 h ahead). Remarkably, zero-shot foundation models demonstrate the ability to outperform trained classical models in data-scarce scenarios, challenging the traditional paradigm of localized training. These findings indicate that optimal forecasting strategies are context-dependent and challenge the assumption of universal model superiority.A.J. and D.P. were partially supported by the Spanish Ministry of Science and Innovation through the project PID2022-136627NB-I00 (MCIN/AEI/10.13039/501100011033/FEDER, UE). P.G.-P. was supported by the Government of Navarra under 'Doctorados Industriales 2021'.MDPIEstadística, Informática y MatemáticasEstatistika, Informatika eta MatematikaInstitute of Smart Cities (ISC)Gobierno de Navarra / Nafarroako Gobernua2026info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionapplication/pdfhttps://hdl.handle.net/2454/57128reponame:Academica-e. Repositorio Institucional de la Universidad Pública de Navarrainstname:Universidad San Jorge (USJ)Inglésinfo:eu-repo/grantAgreement/AEI/Plan Estatal de Investigación Científica y Técnica y de Innovación 2021-2023/PID2022-136627NB-I00© 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license.https://creativecommons.org/licenses/by/4.0/info:eu-repo/semantics/openAccessoai:dnet:academicae__::0f662a392a2ef08611d8cc4bea5342c72026-06-17T12:41:47Z
dc.title.none.fl_str_mv The impact of occupancy dynamics on indoor CO2 forecasting: a cross-scenario evaluation
title The impact of occupancy dynamics on indoor CO2 forecasting: a cross-scenario evaluation
spellingShingle The impact of occupancy dynamics on indoor CO2 forecasting: a cross-scenario evaluation
Garcia-Pinilla, Peio
Indoor air quality
Forecasting
Machine learning
Foundation models
Indoor CO2
IoT
title_short The impact of occupancy dynamics on indoor CO2 forecasting: a cross-scenario evaluation
title_full The impact of occupancy dynamics on indoor CO2 forecasting: a cross-scenario evaluation
title_fullStr The impact of occupancy dynamics on indoor CO2 forecasting: a cross-scenario evaluation
title_full_unstemmed The impact of occupancy dynamics on indoor CO2 forecasting: a cross-scenario evaluation
title_sort The impact of occupancy dynamics on indoor CO2 forecasting: a cross-scenario evaluation
dc.creator.none.fl_str_mv Garcia-Pinilla, Peio
Jurío Munárriz, Aránzazu
Figols, María
Paternain Dallo, Daniel
author Garcia-Pinilla, Peio
author_facet Garcia-Pinilla, Peio
Jurío Munárriz, Aránzazu
Figols, María
Paternain Dallo, Daniel
author_role author
author2 Jurío Munárriz, Aránzazu
Figols, María
Paternain Dallo, Daniel
author2_role author
author
author
dc.contributor.none.fl_str_mv Estadística, Informática y Matemáticas
Estatistika, Informatika eta Matematika
Institute of Smart Cities (ISC)
Gobierno de Navarra / Nafarroako Gobernua
dc.subject.none.fl_str_mv Indoor air quality
Forecasting
Machine learning
Foundation models
Indoor CO2
IoT
topic Indoor air quality
Forecasting
Machine learning
Foundation models
Indoor CO2
IoT
description Indoor CO2 forecasting supports proactive ventilation control that balances air quality with energy efficiency. While Machine Learning (ML) models have shown strong performance in controlled settings such as schools, their generalization across indoor spaces with diverse occupancy dynamics remains poorly characterized. We present a systematic benchmark of 11 forecasting models spanning simple baselines, statistical methods, classical ML, deep learning, ensembles, and foundation models using 18 weeks of IoT sensor data spanning six real-world use cases: conference rooms, dining halls, hospitals, food markets, offices and student residences. Performance depends strongly on the prediction horizon and on the regularity of occupancy-driven CO2 patterns. Simple baselines tend to perform best at short horizons (10 min ahead), while ensembles and fine-tuned foundation models provide more robust accuracy at longer horizons (4 h ahead). Remarkably, zero-shot foundation models demonstrate the ability to outperform trained classical models in data-scarce scenarios, challenging the traditional paradigm of localized training. These findings indicate that optimal forecasting strategies are context-dependent and challenge the assumption of universal model superiority.
publishDate 2026
dc.date.none.fl_str_mv 2026
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv https://hdl.handle.net/2454/57128
url https://hdl.handle.net/2454/57128
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.relation.none.fl_str_mv info:eu-repo/grantAgreement/AEI/Plan Estatal de Investigación Científica y Técnica y de Innovación 2021-2023/PID2022-136627NB-I00
dc.rights.none.fl_str_mv https://creativecommons.org/licenses/by/4.0/
info:eu-repo/semantics/openAccess
rights_invalid_str_mv https://creativecommons.org/licenses/by/4.0/
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv MDPI
publisher.none.fl_str_mv MDPI
dc.source.none.fl_str_mv reponame:Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
instname:Universidad San Jorge (USJ)
instname_str Universidad San Jorge (USJ)
reponame_str Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
collection Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869416389578588160
score 15,811543

Similares en LA Referencia