The news
Columbus-based Path Robotics has landed a $56M for its autonomous welding robots. The system uses the usual combination of scanning technology, computer vision, and AI to adjust itself automatically, and optimize the welding process for each individual part.
This news is buried in most coverage, but these autonomous welding systems are also turnkey. Path Robotics says the systems require no complex robotics programming, even in “high mix, high maintenance, low volume” environments.
Path offers the system with a subscription model, and includes preventative maintenance.
Why it matters
Welding is an essential part of the manufacturing process, and one that requires skilled human labor. The first problem with that, as Path notes in its press release, is that skilled welders are a vanishing resource. The American Welding Society projects a shortage of 400,000 by the end of 2024, even as demand increases steadily.
In response, Path Robotics is using CV, AI, and robotics to offer a system that can weld as well—if not better than—human welders.
Crucially, Path is not requiring that manufacturers replace one type of skilled labor with another. Since the system doesn’t require programming to work with different kinds of parts, and Path offers maintenance, that means manufacturers won’t have to add an expensive roboticist to their staff to take advantage of automation.
Path argues that the system could help companies meet the need for skilled welding, while reducing the need for re-work. In turn, the system could offer a way for companies to “reshore” their manufacturing, offer more competitive salaries for other workers, and scale capacity.
How it works
Details are thin, with not many specifics to be found in Path Robotics’ press release or their slick website. But here are the basics.
The system uses a proprietary scanning technology to create a 3D model of each part individually. Then, it uses CV to identify the seams that need welding. As it begins to weld, it determines the best possible weld path, and re-positions the robot and part to perform that weld.Meanwhile, the system collects data to make adjustments on the fly. This, in turn, allows the system to improve its welding over time.
For more information, see Path Robotics.