Part quality in PBF-LB/M strongly depends on the precise application of metal powder. A typical recoating method is the use of a recoater blade. Damaged or misaligned blades often lead to an inhomogeneous powder layer thickness, local powder layer interruptions, or powder accumulations, which can cause critical defects in industrial application. As a result, lack-of-fusion defects and geometric deviations can diminish the part quality and process stability. This study aims to detect blade misalignment and recoater crashes on an industrial PBF-LB/M machine (EOS M300-4) by analysing data derived from in-situ process monitoring (ISPM) and the machine control system. In particular these are melt pool emissions, measured with a commercial melt pool monitoring (MPM) system, the images of a powder bed camera and recoater torque values, derived from the recoater motor current. The MPM signals are analysed in terms of their part- and layer-wise statistical distribution. The mean value within a part layer shows a good correlation with coating quality. Regions with insufficient powder distribution show a lower mean value compared to nominal coated regions. Peaks of the recoater torque signal due to blade collisions are compared with camera images of the powder bed to evaluate a relation. Based on these findings, practical recommendations for sensor signal thresholds are provided to enable a robust detection of recoating defects.