IROS: A Dual-Process Architecture for Real-Time VLM-Based Indoor Navigation

TL;DR

IROS is a real-time indoor navigation system that combines fast reflexive responses with deliberate reasoning using lightweight vision-language models, enabling robust and efficient robot navigation in complex environments.

Contribution

The paper introduces a dual-process architecture for indoor navigation that integrates VLMs with perceptual modules, reducing latency and improving decision accuracy on embedded hardware.

Findings

Reduces navigation latency by 66% compared to continuous VLM use.

Enhances decision accuracy in real-world indoor environments.

Operates efficiently on low-cost, on-device hardware.

Abstract

Indoor mobile robot navigation requires fast responsiveness and robust semantic understanding, yet existing methods struggle to provide both. Classical geometric approaches such as SLAM offer reliable localization but depend on detailed maps and cannot interpret human-targeted cues (e.g., signs, room numbers) essential for indoor reasoning. Vision-Language-Action (VLA) models introduce semantic grounding but remain strictly reactive, basing decisions only on visible frames and failing to anticipate unseen intersections or reason about distant textual cues. Vision-Language Models (VLMs) provide richer contextual inference but suffer from high computational latency, making them unsuitable for real-time operation on embedded platforms. In this work, we present IROS, a real-time navigation framework that combines VLM-level contextual reasoning with the efficiency of lightweight perceptual modules on low-cost, on-device hardware. Inspired by Dual Process Theory, IROS separates fast reflexive decisions (System One) from slow deliberative reasoning (System Two), invoking the VLM only when necessary. Furthermore, by augmenting compact VLMs with spatial and textual cues, IROS delivers robust, human-like navigation with minimal latency. Across five real-world buildings, IROS improves decision accuracy and reduces latency by 66% compared to continuous VLM-based navigation.