Autonomous Navigation in Mediterranean Greenhouses Using ROS 2-based Reinforcement Learning
DOI:
https://doi.org/10.17979/ja-cea.2026.47.13719Keywords:
Robotics in Agriculture, Mobile Robotics, Reinforcement Learning, PPO, Artificial IntelligenceAbstract
Automation in protected agriculture is key for optimising productivity. However, the autonomous navigation of robots in Mediterranean greenhouses poses severe challenges due to geometric constraints, narrow aisles and dense foliage, where traditional algorithms prove insufficient. To address this problem, this work proposes an autonomous navigation system based on ROS 2 that integrates a reinforcement learning agent based on the Proximal Policy Optimization (PPO) algorithm on the AgriCobIoT I differential mobile robot. The main contribution lies in the design and adaptation of a reward function specifically modelled to mitigate the severe conditions of the agricultural environment and guide learning in restrictive scenarios. The results demonstrate that the proposed strategy enables the training and validation of reinforcement learning agents before their transfer to the real environment.
References
Almeida, F., Leão, G., Sousa, A., 2023. An educational kit for simulated robot learning in ros 2. In: Iberian Robotics conference. Springer, pp. 513–525.
Cañadas-Aranega, F., Gil, J. D., Moreno, J. C., Blanco-Claraco, J. L., 2026. Marco metodológico para la navegación de robots en invernaderos mediante aprendizaje por refuerzo. Simposios del Comité Español de Automática (CEA) 2 (2).
Cañadas-Aránega, F., Mañas-Álvarez, F. J., Moreno, J. C., Blanco-Claraco, J. L., et al., 2026a. A ros2 benchmarking framework for hierarchical control strategies in mobile robots for mediterranean greenhouses. arXiv preprint arXiv:2602.15162.
Cañadas-Aránega, F., Moreno, J. C., Blanco-Claraco, J. L., 2026b. Greenseg: Ground segmentation algorithm for agricultural robots in mediterranean greenhouses using rgb-d point clouds. arXiv preprint arXiv:2605.25279.
Cañadas-Aránega, F., Wollherr, D., Guzmán, J. L., Moreno, J. C., Blanco, J. L., August 22-28 2026c. An adaptive control architecture for slope and terrain compensation in autonomous navigation in mediterranean greenhouses. In: Proceedings of the 23rd IFAC World Congress. Busan, South Korea, accepted for publication.
Cañadas-Aránega, F., Blanco-Claraco, J. L., Moreno, J. C., Rodriguez-Diaz, F., 2024. Multimodal mobile robotic dataset for a typical mediterranean greenhouse: The greenbot dataset. Sensors 24 (6).
Gao, Y., Jiang, Q., Wang, M., Dong, X., 2026. Advances in path-planning algorithms for agricultural robots. Journal of Field Robotics 43 (1), 89–145.
Gil, J. D., Chanona, E. A. D. R., Guzmán, J. L., Berenguel, M., 2026. Reinforcement learning meets bioprocess control through behavior cloning: Real-world deployment in an industrial photobioreactor. Engineering Applications of Artificial Intelligence 164, 113326.
Hernández, H. A., Mondragón, I. F., González, S. R., Pedraza, L. F., 2025. Reconfigurable agricultural robotics: Control strategies, communication, and applications. Computers and Electronics in Agriculture 234, 110161.
Kim, C. H., Silwal, A., Kantor, G., 2025. Autonomous robotic pepper harvesting: Imitation learning in unstructured agricultural environments. IEEE Robotics and Automation Letters.
Lai, M., Go, K., Li, Z., Kröger, T., Schaal, S., Allen, K., Scholz, J., 2025. Roboballet: Planning for multirobot reaching with graph neural networks and reinforcement learning. Science Robotics 10 (106), eads1204.
López-Gázquez, A., Mañas-Álvarez, F. J., Moreno, J. C., Cañadas-Aránega, F., Sánchez, J. A., October 22–27 2023. Navigation of a differential robot for transporting tasks in mediterranean greenhouses. In: Proceedings of the 2023 International Symposium on New Technologies for Sustainable Greenhouse Systems (Greensys). ISHS (International Society for Horticultural Science), Canc´un, Mexico, pp. 1–8.
Macenski, S., Foote, T., Gerkey, B., Lalancette, C., Woodall, W., 2022. Robot operating system 2: Design, architecture, and uses in the wild. Science Robotics 7 (66), eabm6074.
Sánchez-Molina, J. A., Rodríguez, F., Moreno, J. C., Sánchez-Hermosilla, J., Giménez, A., 2024. Robotics in greenhouses. scoping review. Computers and Electronics in Agriculture 219, 108750.
Schulman, J., Levine, S., Abbeel, P., Jordan, M., Moritz, P., 2015. Trust región policy optimization. In: International conference on machine learning. PMLR, pp. 1889–1897.
Schulman, J., Wolski, F., Dhariwal, P., Radford, A., Klimov, O., 2017. Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347.
Wang, P., He, P., Ma, C., Niu, C., Gao, H., Wang, H., Muyeen, S., Zhou, D., 2025. A novel path planning approach for plant protection uav based on ddpg and ila optimization algorithm. Computers and Electronics in Agriculture 239, 111006.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Fernando Cañadas-Aránega, Juan D. Gil, José C. Moreno
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
