Do you have these challenges?

This seminar explains a complete methodology to visualize the “invisible enemy” hidden in vibration problems using data and AI, enabling practical control in real-world applications.
We systematically explain how to acquire data, extract features, and detect anomalies using AI for non-stationary and nonlinear phenomena that cannot be captured by conventional FFT or FEM methods.
We also focus on practical methods that can be immediately applied in the field, such as Excel-based preprocessing and data reconstruction.

・ The nature and mechanism of the “invisible enemy”
・ Impact of measurement conditions and data quality
・ AI-ready data design and feature extraction
・ Selection and application conditions of anomaly detection algorithms
・ Case studies of anomaly detection and cause estimation using field data

• 1. Fundamentals of Vibration Problems and the “Invisible Enemy”
• 
  1-1 Why vibration problems often remain “unidentified”
  1-2 Hidden factors: time variation, non-stationarity, and nonlinearity
  1-3 What is overlooked by conventional analysis (FFT, FEM)
  1-4 Defining the “invisible enemy”: gaps in phenomenon, data, and interpretation
  
  2. Key Points in Vibration Data Acquisition (Preventing Loss of the Invisible Enemy)
• 
  2-1 Measurement location design
  2-2 Sampling conditions and the essence of aliasing
  2-3 Differences between stationary and non-stationary data acquisition
  2-4 Sensor selection and mounting errors causing “false phenomena”
  2-5 Critical impact of missing data and noise on AI
• 
  3. Differences Between Usable and Unusable Data for AI
  
  3-1 What is “unusable data”? (insufficient information, bias)
  3-2 Pitfalls of training data (ambiguous definitions of normal/abnormal)
  3-3 Strategies for unlabeled data utilization
  3-4 Designing systems that work with small datasets
  3-5 Handling variability in field data (treatment of residuals)
  
  4. Fundamentals of Feature Engineering (Quantifying the Invisible Enemy)

• 5. Excel-Based Data Preprocessing and Reconstruction (Practical Implementation)

• 6. Basics of Anomaly Detection Algorithms and Selection Methods
• 7. Case Studies of Anomaly Detection Using Field Data (Visualizing the Invisible Enemy)
• 8. Practical Implementation (Building a System to Control the Invisible Enemy)
• 9. Q&A