LTLCodeGen: Code Generation of Syntactically Correct Temporal Logic for Robot Task Planning
Abstract
This paper focuses on planning robot navigation tasks from natural language specifications. We develop a modular approach, where a large language model (LLM) translates the natural language instructions into a linear temporal logic (LTL) formula with propositions defined by object classes in a semantic occupancy map. The LTL formula and the semantic occupancy map are provided to a motion planning algorithm to generate a collision-free robot path that satisfies the natural language instructions. Our main contribution is LTLCodeGen, a method to translate natural language to syntactically correct LTL using code generation. We demonstrate the complete task planning method in real-world experiments involving human speech to provide navigation instructions to a mobile robot. We also thoroughly evaluate our approach in simulated and real-world experiments in comparison to end-to-end LLM task planning and state-of-the-art LLM-to-LTL translation methods.
Keywords
Cite
@article{arxiv.2503.07902,
title = {LTLCodeGen: Code Generation of Syntactically Correct Temporal Logic for Robot Task Planning},
author = {Behrad Rabiei and Mahesh Kumar A. R. and Zhirui Dai and Surya L. S. R. Pilla and Qiyue Dong and Nikolay Atanasov},
journal= {arXiv preprint arXiv:2503.07902},
year = {2025}
}
Comments
Accepted to IEEE/RSJ International Conference on Intelligent Robots and Systems (iROS) 2025