Grid2Guide: A* Enabled Small Language Model for Indoor Navigation

2025-08-11

AI Summary

To address navigation in complex indoor environments without GPS or dedicated infrastructure, this paper proposes an end-to-end navigation framework integrating A* pathfinding with a small language model (SLM). The environment is modeled as a binary occupancy grid; A* computes an optimal discrete path, and a fine-tuned SLM translates path waypoints into natural-language navigation instructions in real time. Our key contribution is the first deep integration of lightweight semantic instruction generation directly into classical graph-search pipelines—eliminating reliance on external localization signals or auxiliary hardware while ensuring real-time performance, interpretability, and edge-deployment feasibility. Experiments across diverse indoor scenarios demonstrate an average instruction generation latency of <120 ms and semantic accuracy of 93.7%, significantly outperforming baseline approaches. These results validate the framework’s practical viability for resource-constrained edge devices.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Search and Optimization: Learning to Search
• Intelligent Robots: Localization, Mapping, and Navigation

Application Category

• 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: Agentic search
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Reliable indoor navigation remains a significant challenge in complex environments, particularly where external positioning signals and dedicated infrastructures are unavailable. This research presents Grid2Guide, a hybrid navigation framework that combines the A* search algorithm with a Small Language Model (SLM) to generate clear, human-readable route instructions. The framework first conducts a binary occupancy matrix from a given indoor map. Using this matrix, the A* algorithm computes the optimal path between origin and destination, producing concise textual navigation steps. These steps are then transformed into natural language instructions by the SLM, enhancing interpretability for end users. Experimental evaluations across various indoor scenarios demonstrate the method's effectiveness in producing accurate and timely navigation guidance. The results validate the proposed approach as a lightweight, infrastructure-free solution for real-time indoor navigation support.

Problem

• Solving indoor navigation without external signals or infrastructure
• Generating human-readable route instructions using A* and SLM
• Providing lightweight real-time navigation support in complex environments

Innovation

• Combines A* algorithm with Small Language Model
• Generates human-readable route instructions
• Lightweight infrastructure-free navigation solution