Most vibration problems arise not from lack of data, but from data that cannot be effectively used.
In this seminar, based on a fundamental understanding of self-excited vibration, we explain in detail how large volumes of field data should be organized and utilized with AI.
Rather than simply introducing AI methods, the focus is on “what kind of data is usable” and “how to reconstruct it,” providing a highly practical approach directly applicable to real-world engineering.
You will systematically learn the full process from identifying unexplained vibration to anomaly detection.
Do you have these challenges?
- You have large amounts of vibration data, but cannot effectively use it
- Many vibration issues remain unexplained, leading to ad-hoc countermeasures
- FFT analysis works, but the root cause cannot be identified
- You want to use AI but do not know where to start
- You want to introduce anomaly detection, but false detections make it impractical
Seminar Overview
This seminar systematically explains the fundamental understanding of vibration problems centered on self-excited vibration and practical AI-based data analysis methods. Rather than focusing only on theory or AI techniques, it emphasizes the “organization, reconstruction, and utilization” of field data, enabling participants to acquire truly applicable analytical skills.
・ Fundamental understanding and identification of self-excited vibration
・ Typical patterns and diagnostic methods for unexplained vibration
・ Proper interpretation and reconstruction of vibration data
・ Feature extraction concepts suitable for AI
・ Practical processes leading to anomaly detection
Seminar Program
- 1. What is self-excited vibration (fundamental difference from forced vibration)
1-1 Fundamental mechanism of vibration sustained by energy supply
1-2 Differences from forced vibration and typical misinterpretations
- 1-3 Technical requirements for identifying self-excited vibration
2. Why causes become unclear (common pitfalls)
2-1 Difficulty of diagnosis due to “invisible external forces”
2-2 Misconception that frequency match equals root cause identification
3. Signs that suggest self-excited vibration
3-1 Vibration that occurs or persists independent of rotational speed or input
3-2 Behavior where amplitude increases, decreases, or diverges over time
3-3 Nonlinear phenomena that appear or disappear depending on conditions
3-4 Cases where countermeasures have little effect
4. Classification and characteristics of typical self-excited vibration
4-1 Friction-induced (stick-slip, etc.)
4-2 Friction characteristics and transition to negative damping
4-3 Fluid-induced (Kármán vortices, flow-induced vibration, etc.)
4-4 Fluid-structure interaction and oscillation conditions
4-5 Structural / control-induced (negative damping, feedback instability)
4-6 Negative damping and control instability
- 5. Identification from field data
5-1 Interpretation of time waveforms
5-2 Avoiding misdiagnosis from FFT dependence
5-3 Separation of synchronous and asynchronous components
6. Practical process for identifying causes (hypothesis → verification)
6-1 Separation procedure
6-2 Verification through parameter variation
6-3 Reproducibility confirmation
- 7. Effective countermeasure patterns (design, structure, operation)
7-1 Concept of stiffness and mass modification
7-2 Conditions for effective damping addition
7-3 Adjustment of contact, fluid, and operational conditions
- 8. Failure cases of countermeasures and their causes
8-1 Risks of countermeasures without identifying causes
8-2 Limitations of simplistic countermeasures
9. Design guidelines for preventing recurrence
9-1 Risk identification at design stage
9-2 Standardization and checklist development
10. Q&A
Key Outcomes of This Seminar
- You will gain the ability to determine whether vibration is self-excited based on field data and escape from “unknown cause” situations.
- You will be able to reconstruct large vibration datasets into AI-usable forms and apply them consistently from anomaly detection to root-cause identification.
Prerequisite Knowledge
- A basic understanding of undergraduate-level mathematics and mechanics is desirable, but even without it, key concepts and essential points will be explained carefully.
Bonus: Email or Zoom Support
- Free Q&A support regarding seminar content (for 15 days after completion)
- Free technical consulting for vibration-related work issues (for 15 days after completion)
Viewing Period
- Available year-round (on-demand seminar)
- You can view the content for 3 consecutive days at your chosen time.
Please specify your preferred 3-day viewing period in the inquiry field after application.
We will adjust schedules as much as possible and notify you of availability.
Recording Year & Duration
- 2026 edition, approx. 5 hours
Course Fee
- Campaign fee: 28,000 yen (all-inclusive / approx. half of standard technical seminars. Subject to change without notice due to campaign pricing)
List of Participating Companies & Feedback
Instructor
| Title & Name |
Aitop Co., Ltd. Senior Technical Consultant
Certified Noise Control Engineer (Japan Acoustical Society)
Technical Development Award, Acoustical Society of Japan
Former Adjunct Lecturer, Nagoya University Graduate School (lectures in English to international students: 2021–2024)
Hideo Kobayashi
|
| Specialty |
Theory and applied engineering of vibration and noise using AI, and related practical technologies |
| Experience |
Over 30 years of experience as a technical consultant and seminar lecturer, with extensive achievements, including long-term lectures at industrial technology centers across Japan and seminars hosted by Nikkan Kogyo Shimbun. |
*The above seminar program is subject to minor changes without notice.