Linear actuator fault classification with accuracy over 99.9%

Early detection of mechanical faults such as backlash, lack of lubrication, and spalling in electromechanical actuators is critical for ensuring safety and operational efficiency across industries like industrial automation, robotics, and aerospace. Leveraging edge AI technology, we can continuously monitor actuator position error and motor current signatures, employing a trained machine learning model optimized for our MSP microcontrollers to identify these faults in their early stages. This localized, real-time analysis achieves over 99.9% accuracy, shifting maintenance from reactive repairs to proactive condition-based strategies and significantly reducing downtime, operational costs, and extending actuator service life.

This dataset contains raw sensor data from a linear actuator rig operated under different loading conditions and motion profiles, sampled at 25 Hz with three input features:position set point, position error, and motor current.

The task is a 4-class classification problem to detect actuator states: normal, backlash, lack of lubrication, and spalling.

The raw data has been pre-processed in-house into 13 CSV files representing 1 normal   condition and 12 fault conditions at various degradation stages (2 backlash stages, 2 lack of lubrication stages, and 8 spalling stages). Each file contains 60 test combinations covering 2 motion profiles, 3 loading conditions, and 10 repetitions, providing comprehensive data for training fault detection and diagnosis algorithms in linear actuators used in automation and control engineering applications. View open-source dataset.

Accelerate development with advanced optimization and searching algorithms for design, training, and depolying AI models with superior performance. 

CCStudio™ Edge AI Studio gives a start to finish workflow for deploying trained models to embedded targets.For developers seeking deeper customization and control, the MSPM0-SDK offers a comprehensive framework for building and integrating edge AI functionality into your own embeded applications.

The MSP-series MCUs deliver scable performance for executing and accelerating linear actuator fault classification models, along with key SoC features critical to your application.