CANDAS Dataset: A Cooling Fan Sound Dataset with Modeled Disturbances and Controlled Experimental Conditions

• Henrique Lima UNICAP
• Cristofer Silva UPE
• Ricardo Nery UNICAP
• Wilmer Córdoba Camacho UNICAP
• Ricardo Brazileiro Instituto SENAI
• Rodrigo de Paula Monteiro UPE
• Andrea Maria Nogueira Cavalcanti Ribeiro UFPE
• Carmelo José Albanez Bastos-Filho UPE
• Mariela Cerrada Universidad Estatal de Milagro
• Diego Pinheiro UNICAP

DOI: https://doi.org/10.5753/eniac.2025.13864

Resumo

Predictive maintenance cuts economic and safety risks in rotating machinery by leveraging vibration and acoustic data, which machine-learning models convert into intelligent fault detectors. Acoustic signals are especially powerful for early fault detection and cooling fans, with simple rotational dynamics, are convenient proxies for complex rotors. Yet existing fan datasets lack disturbance models and controlled conditions. We present CANDAS, a controlled sound dataset featuring 28 h of recordings from two cooling fans under five modeled disturbance conditions. Baseline experiments with three anomalydetection models validate its value, advancing reproducible research on acoustic fault detection in rotating machinery.

Referências

Abdul, Z. K. and Al-Talabani, A. K. (2022). Mel frequency cepstral coefficient and its applications: A review. IEEE Access, 10:122136–122158.

Al-Hazmi, M. W. (2020a). Cooling fan fault diagnostics using vibrational and acoustical analyses. International Journal of Advanced Science and Technology, 29(3).

Al-Hazmi, M. W. (2020b). Cooling Fan Fault Diagnostics Using Vibrational and Acoustical Analyses. International Journal of Advanced Science and Technology, 29(03).

Alhazmi, M. W., Alsoufi, M. S., Bawazeer, S. A., Hijji, H. H., Alhazmi, H., and Alqurashi, H. F. (2025). Vibration and aeroacoustic analysis of defective cooling fans: effects of blade faults on noise, turbulence, and performance efficiency. Vibration and aeroacoustic analysis of defective cooling fans: effects of blade faults on noise, turbulence, and performance efficiency, pages 1–16.

Bortoni, L. and Jaskowiak, P. (2024). Acoustic features and autoencoders for fault detection in rotating machines: A case study. In Anais da XXXIV Brazilian Conference on Intelligent Systems, pages 34–49, Porto Alegre, RS, Brasil. SBC.

Chandola, V., Banerjee, A., and Kumar, V. (2009). Anomaly detection: A survey. ACM computing surveys (CSUR), 41(3):1–58.

Chinniah, Y. (2015). Analysis and prevention of serious and fatal accidents related to moving parts of machinery. Safety science, 75:163–173.

Dohi, K., Nishida, T., Purohit, H., Tanabe, R., Endo, T., Yamamoto, M., Nikaido, Y., and Kawaguchi, Y. (2022). Mimii dg: Sound dataset for malfunctioning industrial machine investigation and inspection for domain generalization task.

Koizumi, Y., Saito, S., Uematsu, H., Harada, N., and Imoto, K. (2019). Toyadmos: A dataset of miniature-machine operating sounds for anomalous sound detection.

Li, C., Sanchez, R.-V., Zurita, G., Cerrada, M., Cabrera, D., and Vásquez, R. E. (2016). Gearbox fault diagnosis based on deep random forest fusion of acoustic and vibratory signals. Mechanical Systems and Signal Processing, 76-77:283–293.

Li, Y., Yu, R., Shahabi, C., and Liu, Y. (2018). Diffusion convolutional recurrent neural network: Data-driven traffic forecasting.

Lima, H., Silva, C., Nery, R., Camacho, W. C., Brazileiro, R., de Paula Monteiro, R., Ribeiro, A. M. N. C., Bastos-Filho, C. J. A., Cerrada, M., and Pinheiro, D. (2025). IoTDataAtelier/CANDAS · Datasets at Hugging Face — huggingface.co. [link].

Liu, F. T., Ting, K. M., and Zhou, Z.-H. (2008). Isolation forest. In 2008 Eighth IEEE International Conference on Data Mining, pages 413–422. IEEE.

Moura Filho, L. F., de Paula Monteiro, R., Pinheiro, D., Endo, P. T., and Ribeiro, A. M. N. (2023). Forecasting imminent failures in electrical industrial centrifuge using machine learning. In 2023 IEEE Latin American Conference on Computational Intelligence (LA-CCI), pages 1–6. IEEE.

Purohit, H., Tanabe, R., Ichige, K., Endo, T., Nikaido, Y., Suefusa, K., and Kawaguchi, Y. MIMII dataset: Sound dataset for malfunctioning industrial machine investigation and inspection. Number: arXiv:1909.09347.

Purohit, H., Tanabe, R., Ichige, K., Endo, T., Nikaido, Y., Suefusa, K., and Kawaguchi, Y. (2019). Mimii dataset: Sound dataset for malfunctioning industrial machine investigation and inspection.

Roelofs, C. M., Gück, C., and Faulstich, S. (2024). Transfer learning applications for autoencoder-based anomaly detection in wind turbines. Energy and AI, 17:100373.

Roy, R. and George, K. T. (2017). Detecting insurance claims fraud using machine learning techniques. In 2017 International Conference on Circuit ,Power and Computing Technologies (ICCPCT), pages 1–6.

Saufi, S. R., Ahmad, Z. A. B., Leong, M. S., and Lim, M. H. (2019). Challenges and Opportunities of Deep Learning Models for Machinery Fault Detection and Diagnosis: A Review. IEEE Access, 7:122644–122662.

Scalabrini Sampaio, G., Vallim Filho, A. R. D. A., Santos Da Silva, L., and Augusto Da Silva, L. (2019). Prediction of Motor Failure Time Using An Artificial Neural Network. Sensors, 19(19):4342.

Schölkopf, B., Platt, J. C., Shawe-Taylor, J., Smola, A. J., and Williamson, R. C. (2001). Estimating the support of a high-dimensional distribution. Neural Computation, 13(7):1443–1471.

Silva, C., Coelho, M., Barros, U., Ribeiro, A. M. N. C., de Paula Monteiro, R., and Pinheiro, D. (2024). A unified framework for replicability of anomaly detection in multiple time series: Enhancing ganf and ransyncoders pipelines with mfcc feature extraction for acoustic data. In 2024 IEEE Latin American Conference on Computational Intelligence (LA-CCI), pages 1–6.

Tanabe, R., Purohit, H., Dohi, K., Endo, T., Nikaido, Y., Nakamura, T., and Kawaguchi, Y. (2021). Mimii due: Sound dataset for malfunctioning industrial machine investigation and inspection with domain shifts due to changes in operational and environmental conditions.

Yi, W., Choi, J.-W., and Lee, J.-W. (2023). Sound-based drone fault classification using multitask learning.

Zhou, H., Yu, K., Zhang, X., Wu, G., and Yazidi, A. (2022). Contrastive autoencoder for anomaly detection in multivariate time series. Information Sciences, 610:266–280.