Advanced AI

This is an excerpt from an article written by Len Zapalowski in FAB Shop Magazine: August issue pg 55

Since the identification of the “manufacturing skills gap” in the U.S., which predicts 2.1 million unfilled jobs by 2030, many solutions and technological advancements have been launched to help bridge the gap, the cost of which is estimated to total $1 trillion in 2030.

Within the pipe welding industry, advancements in automation continue to provide solutions to the “manufacturing skills gap” caused by an acute shortage of skilled welders, required to meet the demand to build America’s essential infrastructure.

Novarc Technologies Inc., a full-stack robotics company specializing in the design and manufacturing of cobots and computer vision AI for automated welding applications, has been a pioneer in welding automation since 2013, providing a solution to the welding labor shortage.

The company’s flagship product is the Spool Welding Robot (SWR™), the world’s first collaborative welding robot which automates the pipe welding process. The SWR works alongside a human welder, to increase productivity, weld quality and consistency, with the ability to potentially recover capital investment within six to 18 months.

Novarc recently launched a new AI machine learning real-time vision processing system called NovEye™ Autonomy. This new product integrates AI algorithms with advanced robotic controls to deliver the highest quality welds with zero operator intervention, maximizing productivity and eliminating errors.

NovEye Autonomy is the result of years of data collected from Novarc’s fleet of SWRs at customer sites in North America, Europe, Middle East and Australia, and has undergone extensive internal and external testing to ensure all of the features are performing and the weld quality meets ASME standard to ensure unmatched precision and reliability.

NovEye Autonomy is embedded in the control system of the SWR and controls the weld in exactly the same way as a human welder would, allowing the boring, repetitive welds to be delegated to the robot. The autonomous welding process overcomes real-world pipe fitting inconsistencies like HiLo and varying root gaps and can immediately detect and adapt welding parameters to maintain weld integrity.

“NovEye Autonomy is the only AI machine learning real-time vision processing system, that constanty improves welds based on data collection and model enhancement,” says Soroush Karimzadeh, Novarc’s Co-Founder and CEO. “Now, a welder can deliver x-ray quality welds with zero operator intervention, and easily handle various root gaps en route to multi-pass & multi-layer automated welding, due to NovEye Autonomy’s sophisticated AI, and powerful hardware.”

NovEye Autonomy utilizes live-stream video from the camera mounted on the SWR’s torch to analyze the scene and detect welding states. Based on these detected states, the AI adjusts welding and motion parameters in real-time.

“NovEye Autonomy continuously improves through our terabyte-scale library of welding videos, ensuring unmatched precision and reliability,” adds Karimzadeh.

“This accomplishment is the result of years of data collected through machine learning, and we are extremely excited to be able to offer this solution to the welding industry,” adds Karimzadeh.

Read the full article here