Hybrid Artificial Intelligence Strategies for Drone Navigation

2024-10-29

Applied Informatics

AI Summary

To address low path-planning efficiency, high collision risk, and insufficient robustness in target search for autonomous UAV navigation, this paper proposes a state-aware hybrid navigation architecture that integrates Proximal Policy Optimization (PPO)-based deep reinforcement learning with an expert rule engine, enabling flight-state-driven dynamic policy switching. A human-interpretable decision framework and a human-in-the-loop closed-loop evaluation mechanism are further introduced to enhance decision transparency and controllability. The key innovations include: (i) the first real-time collaborative mechanism between RL and rule-based modules; and (ii) an eXplainable AI (XAI)-enhanced navigation paradigm adaptable to diverse operational scenarios. Experimental results demonstrate a 90% success rate in goal-reaching tasks, a significant reduction in collision rate, and a 20% decrease in multi-target search and localization time—collectively validating substantial improvements in both system reliability and operational efficiency.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Search and Optimization: Learning to Search
• Intelligent Robots: State Estimation

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Objective: This paper describes the development of hybrid artificial intelligence strategies for drone navigation. Methods: The navigation module combines a deep learning model with a rule-based engine depending on the agent state. The deep learning model has been trained using reinforcement learning. The rule-based engine uses expert knowledge to deal with specific situations. The navigation module incorporates several strategies to explain the drone decision based on its observation space, and different mechanisms for including human decisions in the navigation process. Finally, this paper proposes an evaluation methodology based on defining several scenarios and analyzing the performance of the different strategies according to metrics adapted to each scenario. Results: Two main navigation problems have been studied. For the first scenario (reaching known targets), it has been possible to obtain a 90% task completion rate, reducing significantly the number of collisions thanks to the rule-based engine. For the second scenario, it has been possible to reduce 20% of the time required to locate all the targets using the reinforcement learning model. Conclusions: Reinforcement learning is a very good strategy to learn policies for drone navigation, but in critical situations, it is necessary to complement it with a rule-based module to increase task success rate.

Problem

• Drone Intelligence
• Autonomous Navigation
• Efficient Pathfinding

Innovation

• Hybrid AI Method
• Reinforcement Learning
• Expert Rules