日本フィジカルAI新聞

世界のフィジカルAIを、日本語で。

歩行arXiv:1906.06502

3Dプリント四足歩行ロボットの設計と材料選定

Designing, 3D Printing of a Quadruped Robot and Choice of Materials for Fabrication

シェア:XThreadsFacebookLINEはてブBluesky

四足歩行ロボットを3Dプリンタで製作し、ABSとアルミ合金の材料比較を通じて脚部の強度と安定性を評価した研究。

著者: Akash Maity, Koustav Roy, Dhrubajyoti Gupta

分類: cs.RO

原文アブストラクト

Purpose-This paper is based on design of a quadruped robot and manufacturing it with 3-D printer followed by its detailed analysis. It focuses on the advantages of additive manufacturing rather than conventional manufacturing techniques and also highlights its limitations. The consequences of choice of different materials for 3-D printing are evaluated in this report. Design/methodology/Approach- The parts were designed in CAD software and made into Stereo lithography files, which were fed into the 3-D printing software. They were printed with the selection of ABS material. Low torque servos were employed in the beginning of the assembly and were controlled with the Arduino Uno micro controller. Findings- The versatility of legged robots require materials with high strength to weight ratio, where 3003 aluminium alloy sheets proved to be a better choice than conventional ABS. The ABS legs buckled under load and proved to be an inferior material choice for fabrication. The aluminium sheet fabricated legs not only imparted structural stability to the robot but also allowed in selection of more powerful servos for added strength.The bot could now achieve significant level of stability. Research Limitations/Implications- The power supply used for powering the bot was of SMPS type of power supply which made it less mobile. Due to the huge demand of current by the high torque servos used later, the power supply became a huge limitation which can be overcome by applying Lithium ion batteries. Practical Implications- The bot can be used for SLAM and autonomous navigation in areas where it is almost impossible for humans to access into. Originality/Value- This paper shows a concrete study on efficacy of quick CAD designing and rapid fabrication.

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、基盤モデル「GO-3 Preview」を発表 身体性AIの3ベンチマークで首位新製品・発表
PR本紙発行元 EmplifAI