Wire arc directed energy deposition robotic additive manufacturing offers near-net-shape fabrication for aerospace components and the repair and modification of cast and forged parts. However, inherent process variability can compromise geometric accuracy and material quality. While in-process sensors have been introduced to mitigate these issues, they often require pausing the process, which limits productivity and adversely affects part quality, especially for large, heavy builds.

This talk will present a real-time, closed-loop control system that continuously adjusts both robot path speed and wire feed rate during the build process. The system integrates in-situ geometry and thermal sensors to enable dynamic monitoring and adaptive control based on a data-driven process model.  This work is performed on a Wire Arc Additive Manufacturing (WAAM) testbed at Rensselaer Polytechnic Institute (RPI).  The system includes a welding robot with a wrist-mounted laser scanner, and a monitoring rob ot equipped with infrared thermal cameras. The system software architecture is based on the Robot Raconteur middleware.