This research focuses on a real-time feedback system to detect and correct defects during the powder bed fusion process of aluminum alloys. In this study, AlSi10Mg coupons were built using various AM parameters. The build process was continuously monitored via a high-frequency in-situ infrared camera which had been integrated into a commercial metal powder bed fusion machine. Porosity information (pore location and size) of the as-built AM coupons were characterized using x-ray computed tomography. The monitoring results were post processed and correlated with porosity location, indicating a strong relationship between abnormal sensing signal and pore formation.