i Weld: an AI-enabled NDT software

Details:

This solution encompasses a real-time AI-enabled image processing platform for the evaluation of weld images captured through non-destructive digital radiography for joint quality assessment including classification and localization of the joint defects.

Copyright:

• AI Enabled NDT System (Diary No: 21253/2023-CO/SW).
• Certified by: Indian Register of Shipping (Certificate no.: 2023PC001)
IRS certificate

Associated industry:

i ToFD: an AI-enabled NDT software

Details:

This solution revolutionizes the detection and sizing of weld defects using ultrasonic time-of-flight diffraction (ToFD) data. This cutting-edge system integrates advanced signal processing with AI algorithms for precise defect localization and measurement.

Patent:

• A method and system for automatically detecting weld defects in ultrasonic Time-of-Flight Diffraction (ToFD) (Filed Ref No: 202331060331). Granted, Patent no,: 529585.
Patent certificate

Associated organization:

Free web-based demo version coming soon...

EyeVib: Insights to Vibration

Details:

This solution provides a machine learning-based system for the health monitoring of the exhauster, which is one of the critical equipment in the steel manufacturing industry.

Solution implemented at: Bokaro Steel Plant

Associated industry:

Free web-based demo version

i TT: intelligent Tyre Tracer

Details:

This AI-driven solution for tyre specification reading and downstream analytics leverages advanced image acquisition, precise text detection, seamless backend integration, and an intuitive user interface.

Associated industry:

Real time weld quality monitoring, prediction and machine control

Details:

This invention proposes a real-time Friction Stir Welding (FSW) process monitoring and control system using multiple sensors. The sensory data is sent to a cloud server for analysis using signal processing and machine learning, enabling real-time quality prediction. When deviations are detected, it adjusts FSW parameters for improved weld quality.

Patents:

• Method and system for multi sensor fusion using tranform learning (Filed Ref No: 202021036163).
• A System for Real Time Monitoring, Prediction and Control of Weld Quality in Friction Stir Welding (Filed Ref No: 202031000072). Granted, Patent no,: 517687.
Patent certificate
• System for real-time monitoring and controlling friction stir welding, (Filed Ref No: 201831024813). Granted, Patent no,: 537304.
Patent certificate

Associated industry:

Real-time asset monitoring

Details:

The solution aims to enhance quality through real-time process monitoring and asset control, thereby leading to significant improvements in product quality. Additionally, it forecasts substantial cost reductions within the Indian work environment, including 20% in factory equipment maintenance costs, 5% in capital equipment investments, 5% in inventory carrying costs, and 15% in transportation expenses. However, the solution also predicts an increase in skilled manpower requirements by 10-20%, aligning with global trends in workforce demands.

Solution implemented at: Tata Metaliks

Associated industry: