Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Reasoning Knowledge-Gap in Drone Planning via LLM-based Active Elicitation

Zeyu Fang, Beomyeol Yu, Cheng Liu, Zeyuan Yang, Rongqian Chen, Yuxin Lin, Mahdi Imani, Tian Lan

TL;DR

This work proposes a novel framework that shifts the collaboration paradigm from control takeover to active information elicitation, and introduces the Minimal Information Neuro-Symbolic Tree (MINT), a reasoning mechanism that explicitly structures knowledge gaps regarding obstacles and goals into a queryable format.

Abstract

Human-AI joint planning in Unmanned Aerial Vehicles (UAVs) typically relies on control handover when facing environmental uncertainties, which is often inefficient and cognitively demanding for non-expert operators. To address this, we propose a novel framework that shifts the collaboration paradigm from control takeover to active information elicitation. We introduce the Minimal Information Neuro-Symbolic Tree (MINT), a reasoning mechanism that explicitly structures knowledge gaps regarding obstacles and goals into a queryable format. By leveraging large language models, our system formulates optimal binary queries to resolve specific ambiguities with minimal human interaction. We demonstrate the efficacy of this approach through a comprehensive workflow integrating a vision-language model for perception, voice interfaces, and a low-level UAV control module in both high-fidelity NVIDIA Isaac simulations and real-world deployments. Experimental results show that our method achieves a significant improvement in the success rate for complex search-and-rescue tasks while significantly reducing the frequency of human interaction compared to exhaustive querying baselines.

Reasoning Knowledge-Gap in Drone Planning via LLM-based Active Elicitation

TL;DR

This work proposes a novel framework that shifts the collaboration paradigm from control takeover to active information elicitation, and introduces the Minimal Information Neuro-Symbolic Tree (MINT), a reasoning mechanism that explicitly structures knowledge gaps regarding obstacles and goals into a queryable format.

Abstract

Human-AI joint planning in Unmanned Aerial Vehicles (UAVs) typically relies on control handover when facing environmental uncertainties, which is often inefficient and cognitively demanding for non-expert operators. To address this, we propose a novel framework that shifts the collaboration paradigm from control takeover to active information elicitation. We introduce the Minimal Information Neuro-Symbolic Tree (MINT), a reasoning mechanism that explicitly structures knowledge gaps regarding obstacles and goals into a queryable format. By leveraging large language models, our system formulates optimal binary queries to resolve specific ambiguities with minimal human interaction. We demonstrate the efficacy of this approach through a comprehensive workflow integrating a vision-language model for perception, voice interfaces, and a low-level UAV control module in both high-fidelity NVIDIA Isaac simulations and real-world deployments. Experimental results show that our method achieves a significant improvement in the success rate for complex search-and-rescue tasks while significantly reducing the frequency of human interaction compared to exhaustive querying baselines.
Paper Structure (18 sections, 1 equation, 2 figures, 1 table)

This paper contains 18 sections, 1 equation, 2 figures, 1 table.

Table of Contents

  1. Introduction
  2. Methodology
  3. Object-Driven Uncertainty Identification
  4. Knowledge-Gap Reasoning
  5. Tree Construction
  6. Evaluation Metrics
  7. Active Elicitation and Plan Refinement
  8. Query Generation
  9. Tree Pruning and Execution
  10. Experimental Results and Discussion
  11. Simulation Experiments: NVIDIA Isaac
  12. Experimental Setup and Baselines
  13. Results Analysis
  14. Real-World Deployment
  15. Scenario Description
  16. ...and 3 more sections

Figures (2)

  • Figure 1: The complete workflow of our methods.
  • Figure 2: The illustration of the testbeds. (a) The typical planning stage in the NVIDIA Isaac simulation environment. (b) The real-world environment for deployment.