Smart Welding: Optimize Friction Stir Welding Using Machine Learning, AI & Digital Twin
- Increase efficiency and productivity through automation.
- Improve quality and reduced defects.
- Get real-time monitoring and predictive maintenance to minimize downtime.
- Get greater customization and flexibility in production.
Overview:
This project introduces an innovative approach to optimizing friction stir welding (FSW) by combining real-time sensor data, machine learning, and digital twin technology. By continuously monitoring key process parameters such as rotational speed, feed rate, and axial force, the system builds predictive models that enable adaptive decision-making and process control. A digital twin provides a dynamic virtual representation of the welding process, allowing for real-time simulation, analysis, and optimization. Experimental trials and AI-based modelling have achieved significant accuracy in predicting defects and optimizing parameters, leading to improved weld quality, reduced material waste, and enhanced overall productivity.
Partners
Sensor Training Data
Enables the models to make accurate predictions or informed decisions based on new data from similar sensors.
Digital Twin
Enables real-time monitoring and serves as a virtual welding machine, assisting in forecasting defects and generating unlimited synthetic training data.
ML / AI App
Identifies relevant features and delivers actionable insights to aid in defect prevention.