AI-powered welding robots for high-volume manufacturing production
Path Robotics builds autonomous welding cells using PyTorch, NVIDIA Isaac Sim, ROS, and embedded Linux to automate repetitive metal-joining tasks. The stack reveals a robotics-first engineering culture: heavy investment in simulation (MuJoCo, Isaac Sim), physics modeling (JAX), and real-time control (ROS 2, gRPC, Protocol Buffers, Jetson), with active work on sim-to-real transfer and whole-body motion control. Hiring has accelerated sharply—37 roles in the last 30 days, heavily weighted toward mid/senior engineers (33 of 37)—signaling rapid scaling of both product development and manufacturing integration.
Path Robotics manufactures autonomous welding systems designed to absorb high-volume, repetitive joining tasks in fabrication shops and contract manufacturers. The company was founded in 2014 by brothers Andy and Alex Lonsberry to address labor shortages in skilled welding roles. The product operates as an integrated cell: customers deploy robotic arms (Yaskawa, ABB) coordinated by Path's control software, guided by computer vision (LiDAR, FARO metrology) and AI-driven path planning. The engineering roadmap emphasizes precision in heavy manufacturing contexts—complex parts, tight tolerances—and addresses core hard problems in the robotics stack: physics simulation accuracy, real-time motion control under constraint, and deployment consistency across diverse shop-floor environments. Path is based in Columbus, Ohio, and operates hiring across North America (US, Mexico, Canada).
Path uses PyTorch, ROS 2, and NVIDIA Isaac Sim for simulation and control. Runtime execution runs on NVIDIA Jetson with embedded Linux, gRPC, and Protocol Buffers. Physical robots are programmed with CODESYS and Studio 5000 (Allen-Bradley) control systems.
Current projects include autonomous welding cell integration into customer facilities, multi-robotic system maintenance, simulator-to-reality transfer for motion tasks, calibration automation, physics-ML hybrid architectures, and sales methodology standardization across GTM.
Other companies in the same industry, closest in size
Path Robotics's technology stack, projects, and hiring signals are inferred from public hiring and company data — career pages, public listings, and company web presence — then clustered and de-duplicated. Figures are estimates that refresh over time. Read our full methodology →
This is not an official vendor or customer list. It is a technology-adoption signal inferred from public data, intended for B2B research.