Learning UAV-based path planning for efficient localization of objects using prior knowledge

2024-12-16

arXiv.org

AI Summary

Agricultural drones suffer from excessive path length and low efficiency when performing row-by-row scanning in fields with non-uniform weed distribution. Method: This paper proposes an adaptive search path planning method based on deep reinforcement learning (DRL). It integrates global prior knowledge with a local object detection map, incorporates an uncertainty-aware mechanism and a dynamic termination strategy, and— for the first time—achieves robust, adaptive search under non-uniform target distributions. Built upon the Deep Q-Network (DQN) framework, the method is trained in simulation and deployed with a lightweight detector, requiring only coarse prior information for stable operation. Results: Experiments demonstrate significantly reduced search time versus conventional row-by-row scanning. Validation spans both simulation and real-world drone tests. The method exhibits strong robustness to detection errors and degradation in prior quality, matching the performance of baseline methods reliant on high-precision priors.

Technology Category

• Search and Optimization: Learning to Search
• Intelligent Robots: Motion and Path Planning
• Planning, Routing, and Scheduling: Learning for Planning and Scheduling

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

UAV's are becoming popular for various object search applications in agriculture, however they usually use time-consuming row-by-row flight paths. This paper presents a deep-reinforcement-learning method for path planning to efficiently localize objects of interest using UAVs with a minimal flight-path length. The method uses some global prior knowledge with uncertain object locations and limited resolution in combination with a local object map created using the output of an object detection network. The search policy could be learned using deep Q-learning. We trained the agent in simulation, allowing thorough evaluation of the object distribution, typical errors in the perception system and prior knowledge, and different stopping criteria. When objects were non-uniformly distributed over the field, the agent found the objects quicker than a row-by-row flight path, showing that it learns to exploit the distribution of objects. Detection errors and quality of prior knowledge had only minor effect on the performance, indicating that the learned search policy was robust to errors in the perception system and did not need detailed prior knowledge. Without prior knowledge, the learned policy was still comparable in performance to a row-by-row flight path. Finally, we demonstrated that it is possible to learn the appropriate moment to end the search task. The applicability of the approach for object search on a real drone was comprehensively discussed and evaluated. Overall, we conclude that the learned search policy increased the efficiency of finding objects using a UAV, and can be applied in real-world conditions when the specified assumptions are met.

Problem

• Efficient UAV path planning for weed localization
• Using reinforcement learning to minimize flight path
• Robust to detection errors and prior knowledge quality

Innovation

• Deep reinforcement learning for UAV path planning
• Combines prior knowledge with in-flight detections
• Robust to detection errors and low-resolution data