日本フィジカルAI新聞

世界のフィジカルAIを、日本語で。

遠隔制御arXiv:2205.04189

FoReCo: 予測ベースの遠隔ロボットマニピュレータ制御復旧機構

FoReCo: a forecast-based recovery mechanism for real-time remote control of robotic manipulators

シェア:XThreadsFacebookLINEはてブBluesky

無線通信の干渉で失われた指令を機械学習で予測・補完し、ロボットマニピュレータの遠隔リアルタイム制御を復旧させる手法を提案・検証した論文。

著者: Milan Groshev, Jorge Martín-Pérez, Carlos Guimarães, Antonio de la Oliva, Carlos J. Bernardos

分類: cs.NI, cs.RO

原文アブストラクト

Wireless communications represent a game changer for future manufacturing plants, enabling flexible production chains as machinery and other components are not restricted to a location by the rigid wired connections on the factory floor. However, the presence of electromagnetic interference in the wireless spectrum may result in packet loss and delay, making it a challenging environment to meet the extreme reliability requirements of industrial applications. In such conditions, achieving real-time remote control, either from the Edge or Cloud, becomes complex. In this paper, we investigate a forecast-based recovery mechanism for real-time remote control of robotic manipulators (FoReCo) that uses Machine Learning (ML) to infer lost commands caused by interference in the wireless channel. FoReCo is evaluated through both simulation and experimentation in interference prone IEEE 802.11 wireless links, and using a commercial research robot that performs pick-and-place tasks. Results show that in case of interference, FoReCo trajectory error is decreased by x18 and x2 times in simulation and experimentation, and that FoReCo is sufficiently lightweight to be deployed in the hardware of already used in existing solutions.

関連論文

  1. 1Boston Dynamics、車輪付きAtlasの研究試作機を紹介研究・技術
  2. 2newmoとKDDIが資本業務提携 2028年に堺市で自動運転タクシー商用化目指す提携・買収
  3. 3GXOとApptronik、Apolloの物流現場試験を拡大 最大1,000台の展開を構想量産・展開
  4. 4NVIDIA、CoRLでGR00T 2とIsaac Lab 3.0を発表新製品・発表
  5. 5UC BerkeleyとNVIDIAの研究者、AIエージェントでロボット制御コードを仮想環境で改善研究・技術

PR・アフィリエイト広告

MINISFORUM(ミニPC)

MINISFORUM(ミニPC)の公式広告バナーMINISFORUM公式ストアを見る ↗
PR本紙発行元 EmplifAI