A Two-Stage Lightweight Framework for Efficient Land-Air Bimodal Robot Autonomous Navigation

2025-07-30

AI Summary

Existing autonomous navigation methods for land–air bimodal robots (LABRs) suffer from two key limitations: insufficient trajectory optimization in map-based approaches and excessive computational overhead in learning-based ones. This paper proposes a two-stage lightweight navigation framework. In the first stage, a keypoint prediction network—integrating edge-aware perception with a lightweight attention mechanism—performs global sparse path planning. In the second stage, local trajectory refinement is achieved via real-time fusion of a Sobel-based perception network and a map-based planner. The framework enables zero-shot simulation-to-reality transfer and runs entirely on CPU. Compared to state-of-the-art methods, it reduces network parameters by 14%, cuts energy consumption during land–air modality switching by 35%, and demonstrates low power consumption, high real-time performance, and strong generalization capability on a custom-built LABR platform.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Planning, Routing, and Scheduling: Planning with Language Models
• Computer Vision: Multi-modal Vision

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search

Abstract

Land-air bimodal robots (LABR) are gaining attention for autonomous navigation, combining high mobility from aerial vehicles with long endurance from ground vehicles. However, existing LABR navigation methods are limited by suboptimal trajectories from mapping-based approaches and the excessive computational demands of learning-based methods. To address this, we propose a two-stage lightweight framework that integrates global key points prediction with local trajectory refinement to generate efficient and reachable trajectories. In the first stage, the Global Key points Prediction Network (GKPN) was used to generate a hybrid land-air keypoint path. The GKPN includes a Sobel Perception Network (SPN) for improved obstacle detection and a Lightweight Attention Planning Network (LAPN) to improves predictive ability by capturing contextual information. In the second stage, the global path is segmented based on predicted key points and refined using a mapping-based planner to create smooth, collision-free trajectories. Experiments conducted on our LABR platform show that our framework reduces network parameters by 14% and energy consumption during land-air transitions by 35% compared to existing approaches. The framework achieves real-time navigation without GPU acceleration and enables zero-shot transfer from simulation to reality during

Problem

• Optimize land-air bimodal robot navigation trajectories
• Reduce computational demands of learning-based methods
• Enable real-time navigation without GPU acceleration

Innovation

• Two-stage lightweight framework for LABR navigation
• Global key points prediction with local refinement
• Sobel Perception Network enhances obstacle detection