Real-time Vision Sensor for Volumetric Flowrate Estimation in Robotic Fused Filament Fabrication

Fused Filament Fabrication (FFF) is one of the most widely used additive manufacturing processes where a part is built layer by layer of deposited polymer material extruded from a heated nozzle. The real-time sensing and control of the FFF process could improve the robustness of the process and the quality of the fabricated part. One of the critical process variables in FFF is the output polymer flow rate, which can be estimated by measuring the thermoplastic material extruded in real-time. Using computer vision, a live video feed of the polymer deposition process can be used as an input for in-situ extrusion width measurements. This paper explores various image processing workflows to improve the measurement latencies of a previously developed vision-based extrusion width measurement approach. The new image processing workflow helps improve the measurement rate by 6x compared to the original workflow, thereby offering prospects of realizing real-time feedback control for FFF.