Achieving Scalable Robot Autonomy via neurosymbolic planning using lightweight local LLM

2025-05-13

AI Summary

Traditional PDDL-based planning suffers from poor scalability, difficulty in real-time replanning, high response latency, and dependence on remote, closed-source large language models—challenging deployment in dynamic human-robot collaboration. Method: We propose Gideon, the first neuro-symbolic planning framework tailored for lightweight, on-device large language models. It integrates the Qwen-2.5-1.5B edge-LLM with symbolic reasoning, employs a generalizable problem generator to construct multi-domain realistic datasets, and leverages context-enhanced prompting and supervised fine-tuning to automate PDDL triplet generation and support long-context reasoning. Contribution/Results: Gideon achieves 66.1% and 70.6% planning success rates on single-domain and multi-domain benchmarks (32k samples), respectively, while reducing model size to just 1/120 of baseline approaches. It significantly improves inference efficiency, deployment flexibility, and cross-domain adaptability—enabling responsive, scalable, and locally executable planning for dynamic human-robot interaction.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Large language models for search

Abstract

PDDL-based symbolic task planning remains pivotal for robot autonomy yet struggles with dynamic human-robot collaboration due to scalability, re-planning demands, and delayed plan availability. Although a few neurosymbolic frameworks have previously leveraged LLMs such as GPT-3 to address these challenges, reliance on closed-source, remote models with limited context introduced critical constraints: third-party dependency, inconsistent response times, restricted plan length and complexity, and multi-domain scalability issues. We present Gideon, a novel framework that enables the transition to modern, smaller, local LLMs with extended context length. Gideon integrates a novel problem generator to systematically generate large-scale datasets of realistic domain-problem-plan tuples for any domain, and adapts neurosymbolic planning for local LLMs, enabling on-device execution and extended context for multi-domain support. Preliminary experiments in single-domain scenarios performed on Qwen-2.5 1.5B and trained on 8k-32k samples, demonstrate a valid plan percentage of 66.1% (32k model) and show that the figure can be further scaled through additional data. Multi-domain tests on 16k samples yield an even higher 70.6% planning validity rate, proving extensibility across domains and signaling that data variety can have a positive effect on learning efficiency. Although long-horizon planning and reduced model size make Gideon training much less efficient than baseline models based on larger LLMs, the results are still significant considering that the trained model is about 120x smaller than baseline and that significant advantages can be achieved in inference efficiency, scalability, and multi-domain adaptability, all critical factors in human-robot collaboration. Training inefficiency can be mitigated by Gideon's streamlined data generation pipeline.

Problem

• Enhancing robot autonomy scalability via local LLMs
• Overcoming dynamic human-robot collaboration limitations
• Enabling multi-domain support with efficient planning

Innovation

• Uses lightweight local LLMs for neurosymbolic planning
• Integrates novel problem generator for scalable datasets
• Enables on-device execution with multi-domain support