日本フィジカルAI新聞

世界のフィジカルAIを、日本語で。

水中ロボットarXiv:2202.01037

RoboKrill:メタクロナル遊泳を模倣したオキアミ型水中ロボット

RoboKrill : a metachronal drag-based swimmer robot

シェア:XThreadsFacebookLINEはてブBluesky

オキアミのメタクロナルな櫂脚運動を再現した小型水中ロボットを設計・製作し、サーボと多リンク機構および流体構造連成で遊泳キネマティクスを実現・検証した。

著者: Sara Oliveira Santos, Francisco Cuenca-Jiménez, P. Antonio Gomez-Valdez, Oscar Morales-Lopez, Monica M. Wilhelmus

分類: cs.RO, physics.bio-ph

原文アブストラクト

Marine exploration is essential to understanding ocean processes and organisms. While the use of current unmanned underwater vehicles has enabled many discoveries, there are still plenty of limitations toward exploring complex environments. Bio-inspired robots are a promising solution for highly maneuverable underwater swimming at moderate speeds. Krill, especially, are efficient swimmers in the intermediate Reynolds number regime and can inform engineering solutions for ocean exploration. In this paper, we present the design, manufacture, and validation of a new krill-inspired, metachronal, drag-based robotic system. By combining active and passive actuation of the joints with 3D printed parts, our unique design recreates the swimming kinematics of Euphausia superba in a compact and reproducible robotic platform. The motion of the anterior and posterior appendage segments is achieved using servo motors and a multi-link mechanism, while the out-of-plane motion of the biramous distal segments is attained via fluid-structure interactions. Going forward, our platform will be leveraged to study metachronal, drag-based swimmers across taxa to identify unifying success mechanisms at different scales, facilitating the development of a new generation of underwater robots.

関連論文

  1. 1パナソニックHDら、ロボット操作AI「T-Rex」を共同開発 12タスクで平均成功率65%研究・技術
  2. 2Hugging Face、ロボット学習基盤「LeRobot v0.7.0」を公開 対話操作と大規模データ学習に対応新製品・発表
  3. 3NVIDIA、Vera CPUのエージェント処理性能を公表 x86比で速度・電力効率を比較半導体・GPU
  4. 4Agility、SPAC合併でナスダック上場資金調達
  5. 5AGIBOTのG2 6台、上海のコンビニで24時間試験 1,730タスクを実施量産・展開
PR本紙発行元 EmplifAI