An investigation on performance of human visual and tactile perception in machined surface inspection

Abstract

Comparative determination of surface roughness by visual and tactile comparison is employed in preliminary part evaluation and quality control in the machining industry. This on-stand evaluation technique allows for quick verification of the product’s adherence with surface roughness specifications. While regarded as less precise than measurement with contact profilometers, it is a fast method that can be employed after basic training. The work conducted by the authors of this study concerned experimental trials under conditions resembling real-life part examination scenarios in industrial settings. Human test subjects were tasked with determining the rougher sample in a pairing in two-alternative forced choice trials, with employed samples taking the form of flat end milled surfaces. Three distinct modes of inspection were involved, namely visual inspection and combined visual and tactile inspection by vertical and horizontal sliding touch, respectively. The obtained results show that the majority of study participants were capable of correctly determining the rougher sample in the pairing, regardless of sensory conditions. Variability in the ratio of correct responses was observed, depending both on stimulus intensity and the employed inspection mode. Overall best performance in terms of ratio of correct responses and variability in responses between participants was noted for combined inspection using vision and vertical sliding touch.