St. Petersburg, Russian Federation
St. Petersburg, Russian Federation
To detect defects and prevent accidents in transport systems, sensors are used that use models based on classical designs or threshold rules, which is fraught with false alarms or delays. Purpose: to propose and describe a practical approach to the use of contrastive self-supervised learning methods for early detection of anomalies in transport monitoring systems. Methods: contrastive learning methods with the InfoNCE loss function, lightweight convolutional encoders (1D-CNN/1D-ResNet), variational autoencoders (VAE) for reconstruction control, attention mechanisms for assessing sensor contributions, adaptive calibration of thresholds based on the exponential moving average (EMA), representation clustering for multiple normal modes, and testing scenarios on semi-real data are applied. Results: features of multisensory time series have been identified, limitations of field devices have been described, and ways to reduce false positives with limited markup have been identified. Algorithmic blocks for practical implementation are proposed: time series augmentation, adaptive threshold calibration, attention mechanisms for explainability and validation on semi-real data. A hybrid anomaly criterion based on contrastive and reconstruction scores has been developed. Practical significance: the introduction of a contrast unit with adaptive calibration provides a decrease in the time for detecting anomalies and a decrease in the frequency of false positives compared to basic VAE and threshold systems. Discussion: integration of contrast modules and adaptive calibration into existing transport monitoring systems using a three-level alarm system with per-channel explanation.
contrastive self-supervised learning, early detection, multi-sensor monitoring, transport monitoring, adaptive calibration
1. Vehicle-as-a-Sensor Approach for Urban Track Anomaly Detection / V. Sruk [et al.] // Sensors. 2025. Vol. 25, iss. 21. Art. 6679. 24 p. DOI:https://doi.org/10.3390/s25216679
2. Thill M., Konen W., Bäck T. Online Anomaly Detection on the Webscope S5 Dataset: A Comparative Study // Proceedings of the 2017 IEEE Workshop on Evolving and Adaptive Intelligent Systems (EAIS 2017) (Ljubljana, Slovenia, 31 May — 2 June 2017). Institute of Electrical and Electronics Engineers, 2017. 8 p. DOI:https://doi.org/10.1109/EAIS.2017.7954844
3. Kingma D. P., Welling M. Auto-Encoding Variational Bayes // Proceedings of the Second International Conference on Learning Representations (ICLR 2014) (Banff, Canada, 14–16 April 2014). 14 p. DOI:https://doi.org/10.48550/arXiv.1312.6114
4. Detecting Spacecraft Anomalies Using LSTMs and Nonparametric Dynamic Thresholding / K. Hundman [et al.] // Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD ‘18) (London, United Kingdom, 19–23 August 2018). New York: Association for Computing Machinery, 2018. Pp. 387–395. DOI:https://doi.org/10.1145/3219819.3219845
5. A Simple Framework for Contrastive Learning of Visual Representations / T. Chen [et al.] // Proceedings of the 37th International Conference on Machine Learning (ICML 2020) (Virtual Event, 13–18 July 2020). Proceedings of Machine Learning Research. 2020. Vol. 119. Pp. 1597–1607.
6. USAD: Unsupervised Anomaly Detection on Multivariate Time Series / J. Audibert [et al.] // Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD‘20), (Virtual Event, 6–10 July 2020). New York: Association for Computing Machinery, 2020. Pp. 3395–3404. DOI:https://doi.org/10.1145/3394486.3403392
7. Deep Autoencoding Gaussian Mixture Model for Unsupervised Anomaly Detection / B. Zong [et al.] // Proceedings of the Sixth International Conference on Learning Representations (ICLR 2018) (Vancouver, Canada, 30 April — 3 May 2018). 19 p. URL: http://openreview.net/forum?id=BJJLHbb0- (data obrascheniya: 09.05.2026).
8. Deng A., Hooi B. Graph Neural Network-Based Anomaly Detection in Multivariate Time Series // Proceedings of the 35th Conference on Artificial Intelligence (AAAI 2021) (Virtual Event, 2–9 February 2021). Proceedings of the AAAI Conference on Artificial Intelligence. 2021. Vol. 35, no. 5. Pp. 4027–4035. DOI:https://doi.org/10.1609/aaai.v35i5.16523
9. On the Evaluation of Unsupervised Outlier Detection: Measures, Datasets, and an Empirical Study / G. O. Campos [et al.] // Data Mining and Knowledge Discovery. 2016. Vol. 30, iss. 4. Pp. 891–927. DOI:https://doi.org/10.1007/s10618-015-0444-8
10. Federated Learning: Challenges, Methods, and Future Directions / T. Li [et al.] // IEEE Signal Processing Magazine. 2020. Vol. 37, no. 3. Pp. 50–60. DOI:https://doi.org/10.1109/MSP.2020.2975749
11. Schneider J. Wenig P., Papenbrock T. Distributed Detection of Sequential Anomalies in Univariate Time Series // The International Journal on Very Large Data Bases. 2021. Vol. 30, iss. 4. Pp. 579–602. DOI:https://doi.org/10.1007/s00778-021-00657-6
12. Long Short Term Memory Networks for Anomaly Detection in Time Series / P. Malhotra [et al.] // Proceedings of the 23rd European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2015) (Bruges, Belgium, 22–24 April 2015). i6doc.com Publishing, 2015. Pp. 89–94.
13. Robust Anomaly Detection for Multivariate Time Series Through Stochastic Recurrent Neural Network / Y. Su [et al.] // Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD ‘19) (Anchorage, AK, USA, 4–8 August 2019). New York: Association for Computing Machinery, 2019. Pp. 2828–2837. DOI:https://doi.org/10.1145/3292500.3330672
