Bio-Inspired Topological Autonomous Navigation with Active Inference in Robotics

2025-08-10

AI Summary

Existing approaches to fully autonomous exploration and navigation in unknown dynamic environments either rely on rigid rule-based systems or require large-scale pretraining data, failing to simultaneously achieve real-time performance, environmental adaptability, and interpretability. Method: This paper proposes a unified probabilistic framework grounded in Active Inference (AIF), integrating Bayesian localization, real-time topological mapping, and adaptive decision-making—enabling end-to-end autonomous navigation without deep learning or pretraining. The framework exhibits strong robustness against dynamic obstacles and pose drift, and is implemented as a modular ROS2 architecture to ensure scalability and engineering deployability. Results: Extensive simulation and real-world experiments demonstrate that our method matches the exploration efficiency and path quality of state-of-the-art planners (e.g., Gbplanner, FAEL), while significantly improving generalization in dynamic environments and providing inherent algorithmic interpretability.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Reasoning under Uncertainty: Probabilistic Inference
• Planning, Routing, and Scheduling: Model-Based Reasoning

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Agentic search

Abstract

Achieving fully autonomous exploration and navigation remains a critical challenge in robotics, requiring integrated solutions for localisation, mapping, decision-making and motion planning. Existing approaches either rely on strict navigation rules lacking adaptability or on pre-training, which requires large datasets. These AI methods are often computationally intensive or based on static assumptions, limiting their adaptability in dynamic or unknown environments. This paper introduces a bio-inspired agent based on the Active Inference Framework (AIF), which unifies mapping, localisation, and adaptive decision-making for autonomous navigation, including exploration and goal-reaching. Our model creates and updates a topological map of the environment in real-time, planning goal-directed trajectories to explore or reach objectives without requiring pre-training. Key contributions include a probabilistic reasoning framework for interpretable navigation, robust adaptability to dynamic changes, and a modular ROS2 architecture compatible with existing navigation systems. Our method was tested in simulated and real-world environments. The agent successfully explores large-scale simulated environments and adapts to dynamic obstacles and drift, proving to be comparable to other exploration strategies such as Gbplanner, FAEL and Frontiers. This approach offers a scalable and transparent approach for navigating complex, unstructured environments.

Problem

• Achieving fully autonomous robot navigation in dynamic environments
• Overcoming reliance on pre-training and large datasets
• Integrating real-time mapping and adaptive decision-making

Innovation

• Bio-inspired Active Inference Framework for navigation
• Real-time topological map creation and updating
• Modular ROS2 architecture for existing systems