AI-powered robotic welding systems for high-volume manufacturing
Path Robotics builds autonomous welding robots designed to handle repetitive, high-volume production tasks in manufacturing environments. The tech stack reveals a hardware-software hybrid company: LabVIEW and Simulink for control systems, Python and SQL for data processing, Kafka and Apache Flink for real-time pipelines, and a data platform built on Snowflake + dbt + Dagster. Active projects around neural welding simulators and physics modeling, paired with pain points in weld parameter selection and physics fidelity, indicate the core challenge is capturing and automating the tacit knowledge embedded in skilled welding operations.
Path Robotics, founded in 2014 by brothers Andy and Alex Lonsberry, addresses workforce gaps in manufacturing by deploying robotic welding systems that absorb high-volume, repetitive tasks. The company targets mid-market fabrication and manufacturing shops facing labor shortages, demand surges, and reshoring pressures. Operations span hardware design and control, software simulation and parameter optimization, and customer integration—reflected in the engineering-heavy hiring mix (nearly half of all open roles) alongside emerging data and product functions. The platform integrates computer vision, motion planning, and real-time feedback to enable autonomous operation on diverse part geometries and weld specifications.
Autonomous robotic welding systems that integrate into customer manufacturing facilities. Core projects include trajectory generation, weld parameter optimization, neural simulators for weld behavior, and scalable data pipelines to reduce cross-team handoffs and rework.
LabVIEW and Simulink for robot control; Python, SQL, Kafka, and Apache Flink for data; Snowflake, dbt, and Dagster for analytics and ETL; Jenkins, GitHub Actions, and Travis CI for CI/CD; Robot Framework and Cucumber for test automation.
Other companies in the same industry, closest in size