About the Lab

ITD Lab (Intelligent Technical Diagnostics) pursues a closed model: from fundamental research in the laboratory to actual industrial products. The Lab develops an AI-based machine health monitoring system — combining signal processing, machine learning/deep learning, and mechanical knowledge — to diagnose failures, predict remaining life (RUL), and optimize maintenance for industrial rotating equipment.

How We Work

1
Background research
Develop vibration signal processing algorithms and AI models (autoencoder, CNN, domain adaptation) for diagnosing camera failures.
2
Verified on real data
Evaluation on standard data sets (CWRU, SEU) and data collected at the factory site, ensuring reliability before deployment.
3
Smart monitoring system
Integrate into a machine health monitoring system: detect abnormalities, diagnose causes, predict RUL and suggest maintenance.
4
Industrial transfer
Training, consulting and implementing solutions at businesses; Aiming for commercial products suitable for Vietnam's industrial conditions.

Our Strengths

Combining AI and mechanical expertise

Unifying artificial intelligence, signal processing and camera dynamics knowledge — a rare advantage compared to data-only solutions.

Hybrid physics-data model

Using digital twin physical platforms combined with machine learning, helps diagnose accurately even when error data is rare.

Explainable AI (XAI)

Apply Grad-CAM, Score-CAM, SHAP to explain model decisions, creating confidence for field engineers.

International cooperation network

Cooperating with MCUT (Taiwan), Griffith (Australia), UTTOP (France), AGH (Poland) and businesses such as VICEM, Samsung.

High quality scientific publications

Many SCIE/Scopus Q1–Q2 papers on vibration diagnostics, deep learning, and camera condition monitoring.

Training linked to practice

Knowledge transfer through corporate training courses and research student teams from 2nd year to PhD.

Three Research Groups

Research students (from year 2 onward) are organised into three main directions.

AI Diagnostics & Prognostics

Diagnose & predict failures using Artificial Intelligence

Development of deep learning and machine learning models for failure diagnosis, anomaly detection and remaining life prediction of rotating equipment: convolutional network (CNN), autoencoder, domain adaptation, explainable AI and data-driven methods for gearbox, bearing, motor monitoring.

Supervisor

PhD Quoc-Chien Truong

Danh-Thanh-Binh Do
- Deep learning-based failure diagnosis for rotating equipment.
- Strong in model building and training.
- Realistic vibration data processing.
Thai-Hung Pham
- Analyze vibration signals and diagnose the rotor–bearing system.
- Strong signal processing foundation.
- Combining time–frequency domain features with AI models.
Manh-Cuong Nguyen
- Data-driven methods for failure diagnosis.
- Proficient in machine learning and feature extraction pipelines.
- Evaluate models, avoid data leaks.
M.Sc. Phuc-Tan Le
- Fellow in AI-based diagnostics.
- Strong in deep learning.
- Build a pipeline to diagnose camera failures.
Nhan-Phuc Hoang
- Additive manufacturing (AM) monitoring.
- Diagnosis in non-stop conditions.
- Data framework for gearbox health monitoring.
Cong-Minh Nguyen
- The cointegration method gives a non-stationary signal.
- Good statistical thinking.
- Applied to anomaly detection.
Dinh-Bach Nguyen
- Process and analyze vibration data for diagnosis.
- Be careful in standardizing data.
- Evaluate the model.
Minh-Thang Bui
- Applying AI to signal diagnosis.
- Eager to learn, quick to use machine learning tools.
- Practice the model.

Bio-Sensor

Biosensors & smart measurement

Research into advanced biosensors and measurement techniques for high-quality signal collection for condition monitoring and diagnostics.

Supervisor

PhD Thi-Mai Tran

Quang-Huy Nguyen
- Development and integration of sensors for signal measurement.
- Strong in data collection system design.
- Sensor signal processing.
Cong-Hieu Le
- Experimenting with sensors and calibrating measurements.
- Meticulous in experimental setup.
- Ensuring the quality of collected data.

Modeling & Simulation

Modeling & simulation of mechatronic systems

Modeling and simulation of rotary transmission system dynamics: gear dynamics, rotor–bearing systems, mechatronic systems and robotics; Building a digital twin physical platform as the basis for physical-data hybrid diagnostics.

Supervisor

PhD Thai-Minh-Tuan Nguyen

Manh-Tuan Le
- Modeling and simulation of the rotor–bearing system.
- Strong in mechanics and dynamic simulation.
- Platform for digital twin.
Tuan-Dong Pham
- Modeling and simulation of gear dynamics.
- Strong in dynamic response analysis.
- Analyze the vibration characteristics of the transmitter.
Van-Anh Hoang
- Modeling and simulation of the rotor–bearing system.
- Combine simulation with real data.
- Calibrate the model.