Localization in vineyards through 3D LiDAR and multi-task neural architecture
DOI:
https://doi.org/10.17979/ja-cea.2026.47.13543Keywords:
Mobile robotics, Autonomous Mobile Robots, agricultural robotics, Pattern recognition and AI in agriculture, Machine Learning, LocalizationAbstract
Obtaining precise location data in an agricultural setting is a complex problem for which no robust solution has been found using state-of-the-art methods. For example, in a vineyard, the repetition of rows makes it difficult to identify distinctive features that would allow for the differentiation of locations in a place recognition task. In this work, we focus on the task of place recognition in vineyard environments using LiDAR sensors. We have extended a convolutional neural network to perform an additional task: classifying the location based on labels that indicate whether the point cloud belongs to a robot turning zone or within a row. We provide a training dataset covering different stages of vine growth, using data from the TEMPO-VINE dataset. The results show that, with a minor modification, the same architecture designed for LiDAR-based location recognition can predict the location label with over 85% precision. These advances pave the way for improving existing LiDAR-based place recognition solutions in agricultural settings.
References
Ali, A., Hussain, T., Zahid, A., 2025. Smart irrigation technologies and prospects for enhancing water use efficiency for sustainable agriculture. AgriEngineering 7 (4), 106. DOI: 10.3390/agriengineering7040106
Allenspach, S., Hiss, J. A., Schneider, G., 2024. Neural multi-task learning in drug design. Nature Machine Intelligence 6 (2), 124–137. DOI: 10.1038/s42256-023-00785-4
Alzahrani, M., Wang, Q., Liao, W., Chen, X., Yu, W., 2024. Survey on multi-task learning in smart transportation. IEEE Access 12, 17023–17044. DOI: 10.1109/ACCESS.2024.3355034
Audrito, G., Martini, M., Navone, A., Galluzzo, G., Chiaberge, M., 2026. VinePT-Map: Pole-Trunk Semantic Mapping for Resilient Autonomous Robotics in Vineyards. arXiv preprint arXiv:2603.05070. DOI: 10.48550/arXiv.2603.05070
Aurelio, Y. S., De Almeida, G. M., de Castro, C. L., Braga, A. P., 2019. Learning from imbalanced data sets with weighted cross-entropy function. Neural processing letters 50 (2), 1937–1949. DOI: 10.1007/s11063-018-09977-1
Barros, T., Garrote, L., Conde, P., Coombes, M., Liu, C., Premebida, C., Nunes, U., 2024. PointNetPGAP-SLC: A 3D LiDAR-Based Place Recognition Approach With Segment-Level Consistency Training for Mobile Robots in Horticulture. IEEE Robotics and Automation Letters. DOI: 10.1109/LRA.2024.3475044
Barros, T., Garrote, L., Conde, P., Coombes, M., Liu, C., Premebida, C., Nunes, U. J., 2023. ORCHNet: A robust global feature aggregation approach for 3d lidar-based place recognition in orchards. arXiv preprint arXiv:2303.00477. DOI: 10.48550/arXiv.2303.00477
Cabrera, J. J., Santo, A., Gil, A., Viegas, C., Pay´a, L., 2025. MinkUNeXt: Point cloud-based large-scale place recognition using 3D sparse convolutions. Array 28, 100569. DOI: 10.1016/j.array.2025.100569
Chen, S., Zhang, Y., Yang, Q., 2024. Multi-task learning in natural language processing: An overview. ACM Computing Surveys 56 (12), 1–32. DOI: 10.1145/3663363
De Silva, R., Swindell, J., Cox, J., Popovi´c, M., Cadena, C., Stachniss, C., Polvara, R., 2025. Keypoint semantic integration for improved feature matching in outdoor agricultural environments. IEEE Robotics and Automation Letters 10 (12), 13383–13390. DOI: 10.1109/LRA.2025.3629991
Eze, V. H. U., Eze, E. C., Alaneme, G. U., BUBU, P. E., Nnadi, E. O. E., Okon, M. B., 2025. Integrating IoT sensors and machine learning for sustainable precision agroecology: enhancing crop resilience and resource efficiency through data-driven strategies, challenges, and future prospects. Discover Agriculture 3 (1), 83. DOI: 10.1007/s44279-025-00247-y
Joseph, T., Fischer, T., Milford, M., 2025. Matched Filtering Based LiDAR Place Recognition for Urban and Natural Environments. IEEE Robotics and Automation Letters. DOI: 10.1109/LRA.2025.3533966
Komorowski, J., 2021. MinkLoc3D: Point cloud based large-scale place recognition. In: Proceedings of the IEEE/CVF winter conference on applications of computer vision. pp. 1790–1799.
Lin, B., Jiang, W., Ye, F., Zhang, Y., Chen, P., Chen, Y.-C., Liu, S., Kwok, J. T., 2023. Dual-balancing for multi-task learning. arXiv preprint arXiv:2308.12029. DOI: 10.48550/arXiv.2308.12029
Lobefaro, L., Sodano, M., Fusaro, D., Magistri, F., Malladi, M. V., Guadagnino, T., Pretto, A., Stachniss, C., 2025. Spatio-temporal consistent semantic mapping for robotics fruit growth monitoring. IEEE Robotics and Automation Letters. DOI: 10.1109/LRA.2025.3594985
Martini, M., Ambrosio, M., Vilella-Cantos, J., Navone, A., Chiaberge, M., 2025. TEMPO-VINE: A Multi-Temporal Sensor Fusion Dataset for Localization and Mapping in Vineyards. arXiv preprint arXiv:2512.04772. DOI: 10.48550/arXiv.2512.04772
Maweni, T., Amayo, P., 2025. Adapting SegMap: A LiDAR place recognition framework for standalone use in C++ applications. In: MATEC Web of Conferences. Vol. 417. EDP Sciences, p. 04004. DOI: 10.1051/matecconf/202541704004
Polvara, R., Molina, S., Hroob, I., Papadimitriou, A., Tsiolis, K., Giakoumis, D., Likothanassis, S., Tzovaras, D., Cielniak, G., Hanheide, M., 2024. Bacchus Long-Term (BLT) data set: Acquisition of the agricultural multimodal BLT data set with automated robot deployment. Journal of Field Robotics 41 (7), 2280–2298. DOI: 10.1002/rob.22228
Salvetti, F., Angarano, S., Martini, M., Cerrato, S., Chiaberge, M., 2022. Waypoint generation in row-based crops with deep learning and contrastive clustering. In: Joint European Conference on Machine Learning and Knowledge Discovery in Databases. Springer, pp. 203–218. DOI: 10.1007/978-3-031-26422-113
Sun, N., Fan, Z., Qiu, Q., Li, T., Feng, Q., Ji, C., Zhao, C., 2024. TriLoc-NetVLAD: Enhancing Long-term Place Recognition in Orchards with a Novel LiDAR-Based Approach. In: 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, pp. 16–22. DOI: 10.1109/IROS58592.2024.10802261
Uy, M. A., Lee, G. H., 2018. PointNetVLAD: Deep point cloud based retrieval for large-scale place recognition. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 4470–4479.
Vilella-Cantos, J., Cabrera, J. J., Pay´a, L., Ballesta, M., Valiente, D., 2025. MinkUNeXt-SI: Improving point cloud-based place recognition including spherical coordinates and LiDAR intensity. arXiv preprint arXiv:2505.17591. DOI: 10.48550/arXiv.2505.17591
Vilella-Cantos, J., Martini, M., Chiaberge, M., Ballesta, M., Valiente, D., 2026. Low cost, high efficiency: Lidar place recognition in vineyards with matryoshka representation learning. Ecological Informatics 95, 103780. DOI: 10.1016/j.ecoinf.2026.103780
Xia, Y., Gladkova, M., Wang, R., Li, Q., Stilla, U., Henriques, J. F., Cremers, D., 2023. CASSPR: Cross attention single scan place recognition. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 8461–8472.
Zheng, Z., Chen, L., Wei, L., Huang, W., Du, D., Qin, G., Yang, Z., Wang, S., 2024. An efficient and lightweight banana detection and localization system based on deep cnns for agricultural robots. Smart Agricultural Technology 9, 100550. DOI: 10.1016/j.atech.2024.100550
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Judith Vilella-Cantos, Míriam Máximo, Óscar Reinoso, Arturo Gil, Mónica Ballesta, David Valiente
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
