CogDDN: A Cognitive Demand-Driven Navigation with Decision Optimization and Dual-Process Thinking

TL;DR

CogDDN is a novel framework for demand-driven robot navigation that mimics human cognition by integrating dual-process decision-making and semantic understanding, leading to improved accuracy and adaptability in unstructured environments.

Contribution

It introduces CogDDN, a VLM-based system combining fast and slow thinking, semantic alignment, and chain of thought reasoning for enhanced demand-driven navigation.

Findings

Outperforms single-view camera methods by 15% in accuracy.

Effectively integrates dual-process decision-making for better adaptability.

Demonstrates robustness in unstructured environments through extensive simulation tests.

Abstract

Mobile robots are increasingly required to navigate and interact within unknown and unstructured environments to meet human demands. Demand-driven navigation (DDN) enables robots to identify and locate objects based on implicit human intent, even when object locations are unknown. However, traditional data-driven DDN methods rely on pre-collected data for model training and decision-making, limiting their generalization capability in unseen scenarios. In this paper, we propose CogDDN, a VLM-based framework that emulates the human cognitive and learning mechanisms by integrating fast and slow thinking systems and selectively identifying key objects essential to fulfilling user demands. CogDDN identifies appropriate target objects by semantically aligning detected objects with the given instructions. Furthermore, it incorporates a dual-process decision-making module, comprising a Heuristic Process for rapid, efficient decisions and an Analytic Process that analyzes past errors, accumulates them in a knowledge base, and continuously improves performance. Chain of Thought (CoT) reasoning strengthens the decision-making process. Extensive closed-loop evaluations on the AI2Thor simulator with the ProcThor dataset show that CogDDN outperforms single-view camera-only methods by 15\%, demonstrating significant improvements in navigation accuracy and adaptability. The project page is available at https://yuehaohuang.github.io/CogDDN/.