A Multi-Output CNN1D–GRU–MLP Architecture for Thermal Time-to-Risk Classification in Cold Chain Logistics
DOI:
https://doi.org/10.61467/2007.1558.2027.v18i1.1465Keywords:
Cold chain logistics, Thermal time series, CNN1D-GRU, Logística de la cadena de frío, series temporales térmicasAbstract
Maintaining an uninterrupted cold chain is essential for ensuring the safety and quality of food and pharmaceutical products. However, conventional monitoring systems mainly provide descriptive information and offer limited support for anticipating thermal risks. This study presents a hybrid CNN1D-GRU-MLP multi-output architecture for thermal time-to-risk classification in multizone cold-chain environments. Due to the limited availability of labeled operational data, the model was initially evaluated using a synthetic dataset generated from energy-balance formulations. Using 30-minute thermal observation windows, the architecture simultaneously performs four tasks: temperature forecasting, time-to-risk interval classification, thermal state classification, and critical zone identification. The interval-bases formulation provides decision-oriented time ranges that are more useful for logistics operations than exact minute-level predictions. Although validation with real-word data remains necessary, the proposes methodology demonstrates the potential to transform thermal trajectories into meaningful information that supports earlier interventions and more effective cold-chain management.
Spanish-language metadata / Metadatos en español
Título en español:
Arquitectura CNN1D–GRU–MLP multisalida para la clasificación del tiempo térmico hasta el riesgo en la logística de la cadena de frío
Resumen:
Mantener una cadena de frío ininterrumpida es esencial para garantizar la seguridad y la calidad de los productos alimentarios y farmacéuticos. Sin embargo, los sistemas convencionales de monitorización proporcionan principalmente información descriptiva y ofrecen un apoyo limitado para anticipar riesgos térmicos. Este estudio presenta una arquitectura híbrida multisalida CNN1D-GRU-MLP para la clasificación del tiempo térmico hasta el riesgo en entornos multizona de cadena de frío.
Debido a la disponibilidad limitada de datos operativos etiquetados, el modelo se evaluó inicialmente utilizando un conjunto de datos sintéticos generado a partir de formulaciones de balance de energía. Mediante ventanas de observación térmica de 30 minutos, la arquitectura realiza simultáneamente cuatro tareas: predicción de temperatura, clasificación del intervalo de tiempo hasta el riesgo, clasificación del estado térmico e identificación de la zona crítica.
La formulación basada en intervalos proporciona rangos temporales orientados a la toma de decisiones que resultan más útiles para las operaciones logísticas que las predicciones exactas a nivel de minutos. Aunque sigue siendo necesaria la validación con datos reales, la metodología propuesta demuestra el potencial para transformar las trayectorias térmicas en información significativa que facilite intervenciones más tempranas y una gestión más eficaz de la cadena de frío.
Palabras Claves:
Logística de la cadena de frío, series temporales térmicas, CNN1D-GRU.
Smart citations:
SciteAI.
Dimensions.
Open Alex.
References
Adefemi, K. O., Mutanga, M. B., & Alimi, O. A. (2025). A hybrid CNN–GRU deep learning model for IoT network intrusion detection. Journal of Sensor and Actuator Networks, 14(5), Article 96. https://doi.org/10.3390/jsan14050096
Artuso, P., Rossetti, A., Minetto, S., Marinetti, S., Moro, L., & Del Col, D. (2019). Dynamic modeling and thermal performance analysis of a refrigerated truck body during operation. International Journal of Refrigeration, 99, 288–299. https://doi.org/10.1016/j.ijrefrig.2018.12.014
Bavaresco, R., Arruda, H., Rocha, E., Barbosa, J., & Li, G.-P. (2021). Internet of Things and occupational well-being in Industry 4.0: A systematic mapping study and taxonomy. Computers & Industrial Engineering, 161, Article 107670. https://doi.org/10.1016/j.cie.2021.107670
Ben Taher, M. A., Kousksou, T., Zeraouli, Y., Ahachad, M., & Mahdaoui, M. (2021). Thermal performance investigation of door opening and closing processes in a refrigerated truck equipped with different phase change materials. Journal of Energy Storage, 42, Article 103097. https://doi.org/10.1016/j.est.2021.103097
Chen, K.-Y., & Shaw, Y.-C. (2011). Applying back propagation network to cold chain temperature monitoring. Advanced Engineering Informatics, 25(1), 11–22. https://doi.org/10.1016/j.aei.2010.05.003
Duan, F., Meng, X., Wu, W., Zou, Y., & Zeng, X. (2026). Data-driven virtual sensor systems for dynamic temperature monitoring along food supply chains. Journal of Stored Products Research, 115, Article 102844. https://doi.org/10.1016/j.jspr.2025.102844
U.S. Food and Drug Administration. (2023). Grade “A” Pasteurized Milk Ordinance (PMO): 2023 revision. U.S. Department of Health and Human Services. FDA official PDF
He, H., & Garcia, E. A. (2009). Learning from imbalanced data. IEEE Transactions on Knowledge and Data Engineering, 21(9), 1263–1284. https://doi.org/10.1109/TKDE.2008.239
Hoang, H. M., Akerma, M., Mellouli, N., Le Montagner, A., Leducq, D., & Delahaye, A. (2021). Development of deep learning artificial neural networks models to predict temperature and power demand variation for demand response application in cold storage. International Journal of Refrigeration, 131, 857–873. https://doi.org/10.1016/j.ijrefrig.2021.07.029
Hu, J., Sari, O., Eicher, S., & Rakotozanakajy, A. R. (2009). Determination of specific heat of milk at different fat content between 1 °C and 59 °C using micro DSC. Journal of Food Engineering, 90(3), 395–399. https://doi.org/10.1016/j.jfoodeng.2008.07.009
International Organization for Standardization. (2009). Milk—Determination of freezing point—Thermistor cryoscope method (Reference method) (ISO Standard No. 5764:2009). ISO official record
Johnson, J. M., & Khoshgoftaar, T. M. (2019). Survey on deep learning with class imbalance. Journal of Big Data, 6, Article 27. https://doi.org/10.1186/s40537-019-0192-5
Khan, S. H., Hayat, M., Bennamoun, M., Sohel, F. A., & Togneri, R. (2018). Cost-sensitive learning of deep feature representations from imbalanced data. IEEE Transactions on Neural Networks and Learning Systems, 29(8), 3573–3587. https://doi.org/10.1109/TNNLS.2017.2732482
KRESS Fahrzeugbau GmbH. (n.d.). Refrigerated truck bodies using an IVECO Daily chassis—Our CoolerBox2.0 has a lot to offer. Retrieved August 7, 2026, from KRESS Iveco Daily page
Maiorino, A., Petruzziello, F., & Aprea, C. (2021). Refrigerated transport: State of the art, technical issues, innovations and challenges for sustainability. Energies, 14(21), Article 7237. https://doi.org/10.3390/en14217237
Mengist, W., Soromessa, T., & Legese, G. (2020). Method for conducting systematic literature review and meta-analysis for environmental science research. MethodsX, 7, Article 100777. https://doi.org/10.1016/j.mex.2019.100777
Mutambik, I. (2026). A blockchain–IoT–ML framework for sustainable vaccine cold chain management in pharmaceutical supply chains. Systems, 14(5), Article 467. https://doi.org/10.3390/systems14050467
Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou, R., Glanville, J., Grimshaw, J. M., Hróbjartsson, A., Lalu, M. M., Li, T., Loder, E. W., Mayo-Wilson, E., McDonald, S., … Moher, D. (2021). The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ, 372, n71. https://doi.org/10.1136/bmj.n71
Page, M. J., Moher, D., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou, R., Glanville, J., Grimshaw, J. M., Hróbjartsson, A., Lalu, M. M., Li, T., Loder, E. W., Mayo-Wilson, E., McDonald, S., … McKenzie, J. E. (2021). PRISMA 2020 explanation and elaboration: Updated guidance and exemplars for reporting systematic reviews. BMJ, 372, n160. https://doi.org/10.1136/bmj.n160
Slaghuis, B. A. (2001). The freezing point of authentic and original farm bulk tank milk in the Netherlands. International Dairy Journal, 11(3), 121–126. https://doi.org/10.1016/S0958-6946(01)00043-7
Thermo King. (n.d.). V-600. Retrieved August 7, 2026, from Thermo King V-600
United Nations Economic Commission for Europe. (2024). ATP handbook 2023: To be used with the Agreement on the International Carriage of Perishable Foodstuffs and on the Special Equipment to be Used for Such Carriage (ATP) as amended on 22 June 2024. United Nations. UNECE ATP Handbook 2023 PDF
United Nations Environment Programme, & Food and Agriculture Organization of the United Nations. (2022). Sustainable food cold chains: Opportunities, challenges and the way forward. https://doi.org/10.4060/cc0923en
Wang, Z., Dong, Y., Liu, W., & Ma, Z. (2020). A novel fault diagnosis approach for chillers based on 1-D convolutional neural network and gated recurrent unit. Sensors, 20(9), Article 2458. https://doi.org/10.3390/s20092458
Xie, Z., Long, H., Ling, C., Zhou, Y., & Luo, Y. (2025). An anomaly detection scheme for data stream in cold chain logistics. PLOS ONE, 20(3), e0315322. https://doi.org/10.1371/journal.pone.0315322
Zhang, K., Huo, X., & Shao, K. (2023). Temperature time series prediction model based on time series decomposition and Bi-LSTM network. Mathematics, 11(9), Article 2060. https://doi.org/10.3390/math11092060
Zhang, Y., Zou, Y., Liu, L., Manzardo, A., Wang, X., & Wu, J. (2026). A digital twin–driven dynamic shelf life approach for mitigating food loss across fruit supply chains. Journal of Stored Products Research, 117, Article 103016. https://doi.org/10.1016/j.jspr.2026.103016
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 International Journal of Combinatorial Optimization Problems and Informatics

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.