Abstract: As carbon fiber reinforced plastics (CFRP) become more integrated into the design of large single piece aircraft structures, a new method was developed which is called Automated Fiber Placement (AFP) to improve the efficiency and quality of CFRP manufacturing. AFP systems allow the repeatable placement of uncured, spool fed carbon fiber tape onto substrates in desired thicknesses and orientations. This automated process may incur defects, such as overlapping, gaps, twists which can severely undermine the structural integrity of the part and decrease its property. Current defect detection and abatement methods are very labor intensive, and still mostly rely on human manual inspection. This part of work is time consuming and vulnerable to human error. Studies have shown that manual inspection can consume more than 20% of total production time. The academia is responding to this by focusing on the development of automated inspection systems. This article describes different inspection methods for automated fiber placement process, including traditional methods, laser projectors, thermal camera, profilometers, machine vision and machine learning. Manual inspection with laser projectors can get higher accuracy but still need manual labor. Thermal camera and machine vision basically identify defects from image which might includes noises. Using profilometer is a promising method because the data from it describes accurate profile of surface. Every method has its own advantages and disadvantages, there are still many major problems need to be solved in AFP defect inspection.

Key words: composites, automated fiber placement, surface defects, inspection technology