Don't Do That!: Guiding Embodied Systems through Large Language Model-based Constraint Generation

Amin Seffo, Aladin Djuhera, Masataro Asai, Holger Boche

Keywords: constrained motion planning, human-aware motion planning, AI-enabled robotics, hallucination mitigation

TL;DR: We introduce a prompting framework to mitigate hallucinations in robot navigation by grounding natural language constraints in verifiable executable predicates.

Abstract: Recent advancements in large language models (LLMs) have spurred interest in robotic navigation that incorporates complex spatial, mathematical, and conditional constraints from natural language into the planning problem. Such constraints can be informal yet highly complex, making it challenging to translate into a formal description that can be passed on to a planning algorithm. In this paper, we propose STPR, a constraint generation framework that uses LLMs to translate constraints (expressed as instructions on "what not to do") into executable Python functions. STPR leverages the LLM's strong coding capabilities to shift the problem description from language into structured and interpretable code, thus circumventing complex reasoning and avoiding potential hallucinations. We show that these LLM-generated functions accurately describe even complex mathematical constraints, and apply them to point cloud representations with traditional search algorithms. Experiments in a simulated Gazebo environment show that STPR ensures full compliance across several constraints and scenarios, while having short runtimes. We also verify that STPR can be used with smaller code LLMs, making it applicable to a wide range of compact models with low inference cost.

Submission Number: 38