What question did this study set out to answer?

The aim is to develop an intelligent screwdriver system that utilizes motion recognition for better assembly precision.

June 3, 2026IET conference proceedings.0 citations

Intelligent screwdriver system with motion recognition using 1D convolutional neural networks

Key Points

The aim is to develop an intelligent screwdriver system that utilizes motion recognition for better assembly precision.
Used a nine-axis MetaMotion C IMU for motion data capture.
Applied Mahony and Kalman filters for calibration and feature extraction.
Employed 1D convolutional neural networks for classifying screwdriver motions.
Achieved 78.3% accuracy in classifying six screwdriver motions, surpassing LSTM baseline accuracy of 66.7%.
Demonstrated improved operational consistency and efficiency in manual assembly tasks.

Abstract

This paper presents an intelligent screwdriver system using one-dimensional convolutional neural networks (1D-CNNs) for real-time motion recognition to enhance assembly precision in industrial automation. A nine-axis MetaMotion C IMU captures motion data, processed with Mahony and Kalman filters for calibration and feature extraction. The 1D-CNN classifies six screwdriver motions with 78.3% accuracy, outperforming the LSTM baseline (66.7%). This scalable solution improves operational consistency and efficiency in manual assembly tasks.

Demander à l'IA

Bookmark

View Full Paper