Optimizing Navigation And Chemical Application in Precision Agriculture With Deep Reinforcement Learning And Conditional Action Tree
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866908282993508352 |
|---|---|
| author | Khosravi, Mahsa Jiang, Zhanhong Waite, Joshua R Jonesc, Sarah Torres, Hernan Singh, Arti Ganapathysubramanian, Baskar Singh, Asheesh Kumar Sarkar, Soumik |
| author_facet | Khosravi, Mahsa Jiang, Zhanhong Waite, Joshua R Jonesc, Sarah Torres, Hernan Singh, Arti Ganapathysubramanian, Baskar Singh, Asheesh Kumar Sarkar, Soumik |
| contents | This paper presents a novel reinforcement learning (RL)-based planning scheme for optimized robotic management of biotic stresses in precision agriculture. The framework employs a hierarchical decision-making structure with conditional action masking, where high-level actions direct the robot's exploration, while low-level actions optimize its navigation and efficient chemical spraying in affected areas. The key objectives of optimization include improving the coverage of infected areas with limited battery power and reducing chemical usage, thus preventing unnecessary spraying of healthy areas of the field. Our numerical experimental results demonstrate that the proposed method, Hierarchical Action Masking Proximal Policy Optimization (HAM-PPO), significantly outperforms baseline practices, such as LawnMower navigation + indiscriminate spraying (Carpet Spray), in terms of yield recovery and resource efficiency. HAM-PPO consistently achieves higher yield recovery percentages and lower chemical costs across a range of infection scenarios. The framework also exhibits robustness to observation noise and generalizability under diverse environmental conditions, adapting to varying infection ranges and spatial distribution patterns. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_17985 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Optimizing Navigation And Chemical Application in Precision Agriculture With Deep Reinforcement Learning And Conditional Action Tree Khosravi, Mahsa Jiang, Zhanhong Waite, Joshua R Jonesc, Sarah Torres, Hernan Singh, Arti Ganapathysubramanian, Baskar Singh, Asheesh Kumar Sarkar, Soumik Robotics Artificial Intelligence This paper presents a novel reinforcement learning (RL)-based planning scheme for optimized robotic management of biotic stresses in precision agriculture. The framework employs a hierarchical decision-making structure with conditional action masking, where high-level actions direct the robot's exploration, while low-level actions optimize its navigation and efficient chemical spraying in affected areas. The key objectives of optimization include improving the coverage of infected areas with limited battery power and reducing chemical usage, thus preventing unnecessary spraying of healthy areas of the field. Our numerical experimental results demonstrate that the proposed method, Hierarchical Action Masking Proximal Policy Optimization (HAM-PPO), significantly outperforms baseline practices, such as LawnMower navigation + indiscriminate spraying (Carpet Spray), in terms of yield recovery and resource efficiency. HAM-PPO consistently achieves higher yield recovery percentages and lower chemical costs across a range of infection scenarios. The framework also exhibits robustness to observation noise and generalizability under diverse environmental conditions, adapting to varying infection ranges and spatial distribution patterns. |
| title | Optimizing Navigation And Chemical Application in Precision Agriculture With Deep Reinforcement Learning And Conditional Action Tree |
| topic | Robotics Artificial Intelligence |
| url | https://arxiv.org/abs/2503.17985 |