Top

GraphSAGE Model with Tether Topology Updates for Underwater Robot Tether Entanglement Prediction

Authors

Files

PDF

Abstract

Underwater robot tether geometry changes as material segments move around obstacles, and future geometric wrapping depends on both recent motion and spatial relationships. This study evaluates a GraphSAGE predictor with permanent material-chain edges and hysteretically updated proximity edges around a vertical cylindrical obstacle. Twenty-four ordered tether nodes carry nine features comprising position, recent velocity, radial clearance, material coordinate and discrete bending. Two neighborhood-aggregation layers and graph-level mean and maximum pooling estimate the probability of a geometric risk state two seconds after a six-frame observation history. The generated benchmark contains 1,500 independent trajectories, partitioned into 900 training, 300 validation and 300 test trajectories. A positive target combines sufficient angular sweep, radial proximity and bounded axial extent at the forecast horizon. Across three training seeds, the updated-topology model attains 94.11% accuracy, 72.65% positive-state recall, F1 0.7534 and Brier score 0.0447. Chain-only and frozen-proximity alternatives attain higher F1 values of 0.7945 and 0.8013, respectively, in this trajectory family. Velocity-omitted retraining yields F1 0.5447; zeroing velocity at inference in a fitted complete model yields F1 0.4863. Proximity connectivity changes over the observation interval in 35.47% of trajectories. Independent position perturbations and graph-delay tests expose sensitivity to descriptor construction and temporal alignment. These results identify predictive value in recent motion while limiting claims of benefit from topology refresh under the declared synthetic geometry. Measured tether motion, contact and tension data are needed to link the geometric label to mechanical entanglement.

PDF

References

Khoshraftar, S., & An, A. (2024). A survey on graph representation learning methods. ACM Transactions on Intelligent Systems and Technology, 15(1), Article 19, 1–55. https://doi.org/10.1145/3633518

Han, J., Cen, J., Wu, L., Li, Z., Kong, X., Jiao, R., Yu, Z., Xu, T., Wu, F., Wang, Z., Xu, H., Wei, Z., Zhao, D., Liu, Y., Rong, Y., & Huang, W. (2025). A survey of geometric graph neural networks: Data structures, models and applications. Frontiers of Computer Science, 19(11), 1911375. https://doi.org/10.1007/s11704-025-41426-w

Zheng, Y., Yi, L., & Wei, Z. (2025). A survey of dynamic graph neural networks. Frontiers of Computer Science, 19(6), 196323. https://doi.org/10.1007/s11704-024-3853-2

Dong, J., Jiang, Z., Pan, D., Chen, Z., Guan, Q., Zhang, H., Gui, G., & Gui, W. (2025). A survey on confidence calibration of deep learning-based classification models under class imbalance data. IEEE Transactions on Neural Networks and Learning Systems, 36(9), 15664–15684. https://doi.org/10.1109/TNNLS.2025.3565159

Yang, L., Chatelain, C., & Adam, S. (2024). Dynamic graph representation learning with neural networks: A survey. IEEE Access, 12, 43460–43484. https://doi.org/10.1109/ACCESS.2024.3378111

Mamedov, S., Geist, A. R., Swevers, J., & Trimpe, S. (2024). Pseudo-rigid body networks: Learning interpretable deformable object dynamics from partial observations. In 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (pp. 9542–9548). IEEE. https://doi.org/10.1109/IROS58592.2024.10802305

Reuel, A., Hardy, A., Smith, C., Lamparth, M., Hardy, M., & Kochenderfer, M. J. (2024). BetterBench: Assessing AI benchmarks, uncovering issues, and establishing best practices. Advances in Neural Information Processing Systems, 37, 21763–21813. https://doi.org/10.52202/079017-0685

Terven, J., Cordova-Esparza, D.-M., Romero-González, J.-A., Ramírez-Pedraza, A., & Chávez-Urbiola, E. A. (2025). A comprehensive survey of loss functions and metrics in deep learning. Artificial Intelligence Review, 58, 195. https://doi.org/10.1007/s10462-025-11198-7

Mamedov, S., Geist, A. R., Viljoen, R., Trimpe, S., & Swevers, J. (2025). Learning deformable linear object dynamics from a single trajectory. IEEE Robotics and Automation Letters, 10(7), 7635–7642. https://doi.org/10.1109/LRA.2025.3577421

Saranti, A., Pfeifer, B., Gollob, C., Stampfer, K., & Holzinger, A. (2024). From 3D point-cloud data to explainable geometric deep learning: State-of-the-art and future challenges. WIREs Data Mining and Knowledge Discovery, 14(6), e1554. https://doi.org/10.1002/widm.1554

Hiniduma, K., Byna, S., & Bez, J. L. (2025). Data readiness for AI: A 360-degree survey. ACM Computing Surveys, 57(9), Article 219, 1–39. https://doi.org/10.1145/3722214

Akasiadis, C., Kladis, E., Kamberi, P.-F., Michelioudakis, E., Alevizos, E., & Artikis, A. (2024). A framework to evaluate early time-series classification algorithms. In Proceedings of the 27th International Conference on Extending Database Technology (pp. 623–635). OpenProceedings.org. https://doi.org/10.48786/edbt.2024.54

Ju, W., Fang, Z., Gu, Y., Liu, Z., Long, Q., Qiao, Z., Qin, Y., Shen, J., Sun, F., Xiao, Z., Yang, J., Yuan, J., Zhao, Y., Wang, Y., Luo, X., & Zhang, M. (2024). A comprehensive survey on deep graph representation learning. Neural Networks, 173, 106207. https://doi.org/10.1016/j.neunet.2024.106207

Feng, Z., Wang, R., Wang, T., Song, M., Wu, S., & He, S. (2026). A comprehensive survey of dynamic graph neural networks: Models, frameworks, benchmarks, experiments and challenges. IEEE Transactions on Knowledge and Data Engineering, 38(1), 26–46. https://doi.org/10.1109/TKDE.2025.3621291

Wang, Y., Hou, W., Sheng, N., Zhao, Z., Liu, J., Huang, L., & Wang, J. (2024). Graph pooling in graph neural networks: Methods and their applications in omics studies. Artificial Intelligence Review, 57, 294. https://doi.org/10.1007/s10462-024-10918-9

Lane, R. O. (2026). A comprehensive review of classifier probability calibration metrics. Data Science, 9(1), Article 24518492261436231. https://doi.org/10.1177/24518492261436231

Wang, A., Tong, T. P., Mizera, A., & Pang, J. (2024). Benchmarking structural inference methods for interacting dynamical systems with synthetic data. Advances in Neural Information Processing Systems, 37, 135129–135185. https://doi.org/10.52202/079017-4295

Zhang, K., Liu, S., Wang, S., Shi, W., Chen, C., Li, P., Li, S., Li, J., & Ding, K. (2026). A survey of deep graph learning under distribution shifts: From graph out-of-distribution generalization to adaptation. ACM Transactions on Knowledge Discovery from Data, 20(2), Article 31, 1–38. https://doi.org/10.1145/3785475

Semmelrock, H., Ross-Hellauer, T., Kopeinik, S., Theiler, D., Haberl, A., Thalmann, S., & Kowald, D. (2025). Reproducibility in machine-learning-based research: Overview, barriers, and drivers. AI Magazine, 46(2), e70002. https://doi.org/10.1002/aaai.70002

Mathieu, T., & Preux, P. (2024). Statistical comparison in empirical computer science with minimal computation usage. In Proceedings of the 2nd ACM Conference on Reproducibility and Replicability (pp. 20–24). ACM. https://doi.org/10.1145/3641525.3663618

Pecher, B., Srba, I., & Bielikova, M. (2024). A survey on stability of learning with limited labelled data and its sensitivity to the effects of randomness. ACM Computing Surveys, 57(1), Article 19, 1–40. https://doi.org/10.1145/3691339

Chan, H. T. J., Šimić, I., & Veas, E. (2025). Explainability and importance estimate of time series classifier via embedded neural network. Scientific Reports, 15, 34515. https://doi.org/10.1038/s41598-025-17703-w

Gravina, A., & Bacciu, D. (2024). Deep learning for dynamic graphs: Models and benchmarks. IEEE Transactions on Neural Networks and Learning Systems, 35(9), 11788–11801. https://doi.org/10.1109/TNNLS.2024.3379735

Zhou, P., Yin, K., Zhu, H., Liang, Y., Li, L., & Chen, Y. (2026). A survey on graph structure learning. Neurocomputing, 696, 134114. https://doi.org/10.1016/j.neucom.2026.134114

Klankers, M. K., & Steil, J. J. (2025). Modeling of deformable linear objects under incomplete state information. In 2025 IEEE International Conference on Robotics and Automation (ICRA) (pp. 10965–10972). IEEE. https://doi.org/10.1109/ICRA55743.2025.11127734

Li, X., Wang, J., & Yan, Z. (2025). Can graph neural networks be adequately explained? A survey. ACM Computing Surveys, 57(5), Article 131, 1–36. https://doi.org/10.1145/3711122

Dai, E., Zhao, T., Zhu, H., Xu, J., Guo, Z., Liu, H., Tang, J., & Wang, S. (2024). A comprehensive survey on trustworthy graph neural networks: Privacy, robustness, fairness, and explainability. Machine Intelligence Research, 21(6), 1011–1061. https://doi.org/10.1007/s11633-024-1510-8

Zhang, W., Zhang, L., Han, J., Liu, H., Fu, Y., Zhou, J., Mei, Y., & Xiong, H. (2024). Irregular traffic time series forecasting based on asynchronous spatio-temporal graph convolutional networks. In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (pp. 4302–4313). ACM. https://doi.org/10.1145/3637528.3671665

Salles, R., Lima, J., Reis, M., Coutinho, R., Pacitti, E., Masseglia, F., Akbarinia, R., Chen, C., Garibaldi, J., Porto, F., & Ogasawara, E. (2024). SoftED: Metrics for soft evaluation of time series event detection. Computers & Industrial Engineering, 198, 110728. https://doi.org/10.1016/j.cie.2024.110728

Ghorbani, R., Reinders, M. J. T., & Tax, D. M. J. (2024). PATE: Proximity-aware time series anomaly evaluation. In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (pp. 872–883). ACM. https://doi.org/10.1145/3637528.3671971

Kose, H. T., Nunez-Yanez, J., Piechocki, R., & Pope, J. (2024). A survey of computationally efficient graph neural networks for reconfigurable systems. Information, 15(7), 377. https://doi.org/10.3390/info15070377

Lu, H., Wang, L., Ma, X., Cheng, J., & Zhou, M. (2025). A survey of graph neural networks and their industrial applications. Neurocomputing, 614, 128761. https://doi.org/10.1016/j.neucom.2024.128761

Details