A record of project delivery by MOORIM T.C.
Vehicle acceleration, facial data, and passenger response signals were collected in time synchronization and structured into a dataset for model training and validation.
The trained model was deployed to a device so inference and data logging could run simultaneously in the field.
Time-series acceleration sensor data was collected, and issues were diagnosed through frequency-domain (FFT) and time-series analysis.
The analysis results were used to quantify ride comfort levels and derive directions for improvement.
A pipeline was built to estimate 6DoF pose and displacement using markers and an industrial camera, generating deformation data for the measurement target.
The system was optimized for convenient use in the field.
Vehicle and passenger state signals such as facial and biometric inputs were collected by camera during driving and accumulated into a dataset.
The input structure and pipeline were organized so the collected data could be used to train and validate a stress estimation model.
Load and displacement data were simultaneously collected by combining a mobile device, ArUco markers, and a wireless load measurement device.
It was implemented as a mobile phone and mobile application form for easy use in the field.
Natural frequency and response data for each subsystem were organized into a vibration mode map and converted into comparable indicators.
The analysis results were organized for use in design review and improvement decisions.
Vibration data generated during mechanical system operation is collected with a mobile app and wireless sensors.
The collected data is analyzed and used to diagnose failure states and other system conditions.