火災検知機能を備えた3D環境における自律マルチロボット探査戦略
Autonomous Multi-Robot Exploration Strategies for 3D Environments with Fire Detection Capabilitie
3D環境での建物探査と火災検知に向け、位置推定・マッピング・軌道計画を統合したマルチロボット探査手法を概説し、分散マップ生成やUAV連携などの将来方向を示した。
著者: Ankit Shaw
分類: cs.RO
原文アブストラクト
This paper presents a comprehensive overview of exploration strategies utilized in both 2D and 3D environments, focusing on autonomous multi-robot systems designed for building exploration and fire detection. We explore the limitations of traditional algorithms that rely on prior knowledge and predefined maps, emphasizing the challenges faced when environments undergo changes that invalidate these maps. Our modular approach integrates localization, mapping, and trajectory planning to facilitate effective exploration using an OctoMap framework generated from point cloud data. The exploration strategy incorporates obstacle avoidance through potential fields, ensuring safe navigation in dynamic settings. Additionally, I propose future research directions, including decentralized map creation, coordinated exploration among unmanned aerial vehicles (UAVs), and adaptations to time-varying environments. This work serves as a foundation for advancing coordinated multi-robot exploration algorithms, enhancing their applicability in real-world scenarios.
