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

2026-03-08

AI Summary

This work addresses the inefficiency and high cognitive load experienced by non-expert operators in cooperative UAV planning due to frequent interventions necessitated by environmental uncertainty. To mitigate this, the authors propose the Minimal Information Neural Symbolic Tree (MINT) mechanism, which leverages a large language model to structurally model knowledge gaps and generate optimal binary queries that actively resolve task-relevant ambiguities—such as those concerning obstacles and targets—with minimal human interaction. The system integrates a large language model, a vision-language model, a speech interface, and low-level control modules into an end-to-end human–robot collaborative planning framework. Experimental results in both high-fidelity simulation and real-world search-and-rescue scenarios demonstrate that the approach significantly improves task success rates while substantially reducing the frequency of human–robot interactions.

Technology Category

• Humans and AI: Planning and Decision Support for Human-Machine Teams
• Planning, Routing, and Scheduling: Planning with Language Models
• Multiagent Systems: Multiagent Planning

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Assisted, interactive, and conversational search

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.

Problem

• knowledge gap
• human-AI collaboration
• UAV planning
• environmental uncertainty
• reasoning

Innovation

• active elicitation
• neuro-symbolic reasoning
• large language models
• human-AI collaboration
• UAV planning