A method for the metrological characterization of eye- and head-tracking interfaces for human–machine interaction through eXtended Reality head-mounted displays

Given the increasing interest in novel and powerful Human–Machine Interaction modalities, as well as the role that eXtended Reality (XR) plays in the contemporary societal framework, this paper proposes a metrologically sound method for the experimental characterization of innovative XR-based Human–Machine Interfaces (HMIs). The aim is to thoroughly analyze the interaction between users and the digital content rendered within the XR environment, focusing on two hands-free HMIs, namely eye- and head-tracking. Starting from the acquisition of eye gaze and head pose, the proposed method encompasses the analysis of metrics such as horizontal and vertical offsets, and Euclidean distance with respect to digital reference points. From these metrics, the proposed method yields a novel figure of merit, named Maximum Number, providing insight into the optimal configuration of the XR content to maximize the information transfer. As a case study, but without loss of generalization, the XR head-mounted display Microsoft HoloLens 2 is considered. Experimental findings, derived from a campaign involving 16 subjects, contribute to a deeper understanding of the accuracy and precision in content selection, along with the number of objects that can be accommodated within the XR environment for the developed HMIs. This addresses a critical gap in current knowledge and offers valuable insights, compensating for the lack of information available in technical specifications, paving the way for the development of reliable hands-free applications in contexts with stringent requirements, such as industry or healthcare inspection tasks.