LLM-Driven Self-Refinement for Embodied Drone Task Planning

TL;DR

SRDrone is a self-refining drone task planning system that combines continuous state evaluation and hierarchical Behavior Tree modifications, significantly improving success rates in industrial drone operations.

Contribution

It introduces a continuous state evaluation method and a hierarchical BT modification model for self-refinement in embodied drone task planning, integrating LLM reasoning with physical constraints.

Findings

44.87% improvement in success rate over baselines

Achieved 96.25% success rate in real-world deployment

Effective integration of LLM reasoning with drone planning constraints

Abstract

We introduce SRDrone, a novel system designed for self-refinement task planning in industrial-grade embodied drones. SRDrone incorporates two key technical contributions: First, it employs a continuous state evaluation methodology to robustly and accurately determine task outcomes and provide explanatory feedback. This approach supersedes conventional reliance on single-frame final-state assessment for continuous, dynamic drone operations. Second, SRDrone implements a hierarchical Behavior Tree (BT) modification model. This model integrates multi-level BT plan analysis with a constrained strategy space to enable structured reflective learning from experience. Experimental results demonstrate that SRDrone achieves a 44.87% improvement in Success Rate (SR) over baseline methods. Furthermore, real-world deployment utilizing an experience base optimized through iterative self-refinement attains a 96.25% SR. By embedding adaptive task refinement capabilities within an industrial-grade BT planning framework, SRDrone effectively integrates the general reasoning intelligence of Large Language Models (LLMs) with the stringent physical execution constraints inherent to embodied drones. Code is available at https://github.com/ZXiiiC/SRDrone.