Predictive Maintenance
Predictive Maintenance - Optimizing Equipment Effectiveness through AI (2 days)
Predictive maintenance is a data-driven maintenance strategy that uses data from industrial processes, PLCs and sensors to predict when maintenance is needed for equipment or machinery. This approach is designed to detect potential issues early, before they result in more significant problems that could cause downtime, decreased productivity, or safety hazards.
By predicting when maintenance is needed, predictive maintenance can help organizations reduce costs associated with unscheduled downtime and prevent catastrophic equipment failures. It can also improve equipment reliability, extend equipment lifespan, and improve overall operational efficiency.
In this course you will learn the characteristics of a predictive maintenance strategy compared with strategies such as reactive and preventive maintenance. You will understand the predictive maintenance workflow from acquiring and preprocessing data, feature engineering, model training, model evaluation, model deployment, and model management. Furthermore you will gain insights into root cause analysis of problems in your machinery to identify which parts need to be repaired or replaced.
The theoretical topics of the training will be intensified by hands-on exercises, based on real-world IIoT datasets. By default, all exercises are performed with Python.
Course Outline
- Introduction to Condition Monitoring and Predictive Maintenance
- Data sources and data acquisition
- Data preprocessing for discrete processes
- Data preprocessing for continuous processes
- Feature Engineering – how to create Condition Indicators
- Anomaly Detection Models (unsupervised failure detection)
- Root-Cause-Analysis for automated failure detection
- Visualizing root-causes with Explainable AI
- Remaining Useful Life Prediction Techniques
- How to train Survival Models
- How to train Degradation Models
- How to train Similarity Models
- Model Assessment Techniques
- Model Deployment and Integration
- Alert Management: Setting the best threshold for AI-based Online Condition Monitoring
- Model Management & Model Operations (MLOps)
Who should attend
Managers of maintenance, production plants, production planning, and production operations. Data Scientists who want to apply their capabilities into practice in the field of condition monitoring and predictive maintenance. Consultants whose objective is to put additional value on the latest innovative technology in the smart factory sector.
What you need to succeed
Basic programming skills in Python (e.g. our training courses Python for Smart Factory Analytics or Getting started with Python).
Did not find the training you are looking for? Please feel free to ask for any other Advanced Analytics training.