The article describes the development and ergonomic evaluation of an augmented reality (AR) welding helmet. The system provides an augmented user interface with supporting information relevant to the welding process. The experimental studies focused on hand-eye coordination of welders and nonwelders with two prototypes of the augmented reality welding helmet. The first prototype operated at 16 frames per second, whereas the second, improved version had 20 frames per second. In addition, the hand-eye coordination while wearing the welding helmet with video see-through head-mounted display was compared to a performance with natural vision, without any helmet. Experimental results showed significant influence of helmet and occupation on handeye coordination. Subjective assessment revealed better rating for stereo perception for the system with the higher frame rate, whereas no significant difference in performance was found between the two frame rates.

The paper describes two experiments for investigating the influence of different levels of camera displacement on hand-eye coordination while using a video see-through head-mounted display. During the first experiment 15 camera positions with five levels of height displacement and three levels of depth displacement were compared in four different tasks. Using a two-way ANOVA, the comparison of the calculated performance characteristic values showed significant influence of height displacement on hand-eye coordination. In conclusion, cameras should be placed above or below eye level, but by no more than 35 mm, in order to preserve hand-eye coordination. In the second experiment, a mirror system was used to check hand-eye coordination in an exemplary medical task allowing the cameras to be placed virtually at eye level. A significant decrease in accuracy was found while using the head-mounted display compared to direct view. Finally, the mirror system was compared to the 15 camera positions using the data from the same tasks. Significant differences in performance were found between the mirror system and eye level position, as well as the position slightly below eye level.The results of the experiment provide design recommendations for developers and users of video see-through systems.