Innovation Fall 2026

upfront but have to be managed on the fly," Gumulia said. The integration of these features allowed more complex welds to be automated more comprehensively, addressing a persistent training and quality-control bottleneck in welding. "Something I don't think the average person realizes is just how difficult some of these welds are," said Van Dyck. "Take an open-root, full penetration TIG weld on stainless steel; it can take a welder two to three years of training to get to the point where they can do that reliably at X-ray quality. "With our robot, I was able to perform that kind of weld after just a couple of days learning to use the system. So the value isn't only about speed; it's also about making incredibly demanding welds accessible to people who don't have years and years of experience." Novarc says that automation has also led to speed advances. For some large-diameter stainless pipes, around 76 cm, shops might spend eight to 10 hours welding a single pipe manually. Using a fully autonomous approach, these types of welds can be completed in about an hour. "When you combine that kind of productivity increase with consistently high quality on very demanding welds, it becomes a huge value proposition," said Van Dyck. Automation also mitigates some of the hazardous aspects of welding. "We'll see fewer humans physically holding the torch, which is the dangerous part," said Karimzadeh, "because welding involves long term hazards: exposure to arc light, carcinogenic gases, and repetitive strain injuries from doing the same motions for many years." New approaches to autonomy In May 2026, Novarc moved beyond its pipe welding offerings to add adaptive welding to existing six axis industrial robots. The company believes that its library of pipe welding data is robust enough

Novarc CEO Soroush Karimzadeh, P.Eng., observes their company's flagship Spool Welding Robot (SWR). P hoto : W endy D P hotography

autonomous product offering for Novarc's spool-welding robots. "It took us a few years to develop a software upgrade for SWR that would allow it to run autonomously," said Karimzadeh. "Once the weld is initiated, computer vision translates into commands the robot can execute at any point in time during welding." Autonomy aims to target issues that can come along with the natural variations in the staging process. "The processes before welding are typically manual," Karimzadeh said. "Humans are versatile: they can interpret drawings, move parts together, fit them and tack them

[weld them in place]. That manual work also creates variation in parts and part preparation." Manufacturers have traditionally managed this problem by controlling the environment around the robot, with fixtures holding each component in a precise location and fitters working within narrow tolerances. Novarc focuses on creating "adaptive" systems, or systems that can work well with a variety of starting conditions. "The system has to handle badly cut parts, thermal expansion and cooling shrinkage; conditions that are difficult or impossible to predict

28

Fall 2026

Innovation