AI guides real time welding path with vision

AI maps the torch path in real time. That straightforward line sits at the core of Path Robotics’ latest push to make welding as much about perception as it is about heat and metal. In a conversation captured for The Robot Report Podcast, Path Robotics cofounder and CEO Andy Lonsberry describes a system that orchestrates a welding robot by watching the work through a live camera feed and then steering the torch along an optimal trajectory. The goal is to remove guesswork from the weld and keep a consistent bead even as parts shift, warp, or vary in fit.

Path’s approach centers on applying physical AI to production realities. The Columbus, Ohio, based company frames its technology as a way to translate human intuition into repeatable, robot assisted welding. The AI identifies the torch path and then moves the robot to follow it, using real time vision guidance to maintain the intended trajectory. The result, in theory, is fewer reworks, tighter tolerances, and faster ramp ups in shops where welding remains a bottleneck. The company’s narrative emphasizes practical constraints, jigs and fixtures alone don’t always account for the jiggle of fabricating large parts or the heat distortions that creep in during a long weld.

The broader arc of manufacturing automation, Path’s story underscores a practical truth: the feasibility leap in welding increasingly rides on perception and planning in motion, not just precise torches and stiff fixtures. The emphasis on real time path planning, vision guided control, and on site mobility reflects a design mentality engineers have long urged, systems that adapt to real world variability rather than expecting the world to conform to the robot’s idealized plan. If the momentum holds, shipyards and automotive shops alike will be watching closely how this blend of AI and mobile robotics translates into reliable welds, safer operations, and measurable productivity gains.

From a practitioner’s standpoint, a few realities jump out. First, the real time vision layer is the hard cap on fidelity. In welding, a few millimeters of misalignment can ripple into a defect. Path’s model must continuously reconcile camera input with seam geometry, torch position, and the robot’s own kinematics, all while the part might be moving or clamping might loosen over time. That requires tight calibration, high bandwidth sensing, and robust fault handling to avoid chasing a false seam when glare, smoke, or poor lighting occlude the view. Second, mobility introduces new dynamics. A Spot based welding workflow adds the uncertainty of a moving base, the need to stabilize the welding arc, and the coordination challenge of keeping the torch aligned while the platform itself shifts across a shop floor. Third, integration with shipbuilding workflows means welding teams must trust the AI’s decisions under safety critical constraints, including interlocks, emergency stops, and human supervision during critical welds. Finally, the path to production hinges on repeatable performance across a family of welds, joints, and materials, not just a single ideal seam demonstrated in a lab.