Sonibel Instruments has launched its real-time weld defect detection system, the latest evolution of in-process weld quality. The release introduces a compact sensor mounted onto a welding torch that analyzes the high-fidelity sound of a weld to detect defects as they are occurring, giving welders the option of repairing defects right away.
With Sonibel's technology, welders can repair defects immediately after they occur so they can pass inspections the first time, saving time and money by reducing the back and forth between rewelds and reinspections.
"We built Sonibel to serve the realities of critical weld operations, whether that is in structural, shipbuilding, pipeline, or defense," said Hooman Pirouz, Chief Product Officer of Sonibel Instruments. "As a shipyard reliability engineer, I saw how much time and money was spent on the back and forth between failed weld inspections, rework, and re-inspections."
Sonibel uses machine learning to differentiate between various weld defects in real time. Each weld is different and fabrication environments are highly unpredictable. As a result, developing a system that can pull out meaningful signals from welds among loud background noise is a key priority. The device's use of AI enables the sensor to catch weld defects with robust efficiency.
Defects are then displayed in real time on a screen mounted to the welding unit, allowing the welder to make immediate corrections, rather than waiting for the post-weld inspection.
Sonibel's system is currently being optimized to detect porosity and lack of fusion in semi-automatic FCAW and GMAW welds. Custom integration is possible across other weld processes and defects, including GTAW, SAW, SMAW, and laser welding, both semi-automatic and automated.
For more information contact:
Sophia Millar, CEO
Sonibel Instruments
236-978-1211
sophia@sonibelinstruments.com
www.sonibelinstruments.com