Hierarchical Sampling-based Planner with LTL Constraints and Text Prompting

2025-01-12

AI Summary

Robots struggle to jointly integrate high-level semantic instructions with formal temporal constraints for safe path planning in complex environments. Method: We propose a hierarchical planning framework jointly driven by Linear Temporal Logic (LTL) and natural language. Our approach pioneers direct compilation of natural language instructions into LTL formulas; constructs a hierarchical map abstraction coupled with a transition system model; and enables end-to-end closed-loop planning—from text input to deterministic finite automaton (DFA), high-level BFS-based task planning, and low-level sampling-based navigation (e.g., RRT/PRM). It supports dynamic instruction updates and semantic-physical co-optimization. Results: Experiments across multiple scenarios demonstrate successful execution of complex tasks involving temporal ordering, obstacle avoidance, and multi-goal constraints (e.g., “first patrol Area A, then retrieve an object, and finally return while avoiding obstacles”). The framework significantly improves task fidelity, safety, and interpretability—establishing a novel paradigm for natural language–guided autonomous robot planning.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Intelligent Robots: Motion and Path Planning
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Search and Retrieval-Augmented AI: Natural language understanding for Web search
• 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

Abstract

This project introduces a hierarchical planner integrating Linear Temporal Logic (LTL) constraints with natural language prompting for robot motion planning. The framework decomposes maps into regions, generates directed graphs, and converts them into transition systems for high-level planning. Text instructions are translated into LTL formulas and converted to Deterministic Finite Automata (DFA) for sequential goal-reaching tasks while adhering to safety constraints. High-level plans, derived via Breadth-First Search (BFS), guide low-level planners like Exploring Random Trees (RRT) and Probabilistic Roadmaps (PRM) for obstacle-avoidant navigation along with LTL tasks. The approach demonstrates adaptability to various task complexities, though challenges such as graph construction overhead and suboptimal path generation remain. Future directions include extending to considering terrain conditions and incorporating higher-order dynamics.

Problem

• Intelligent Planner
• Complex Environment
• Path Optimization

Innovation

• Hierarchical Random Selection
• Linear Temporal Logic
• Path Optimization