Monocular Markerless Motion Capture for Concept-Stage Human-Factors Evaluation of an IVD Sample-Loading Unit

Early layout decisions in in vitro diagnostic devices can affect operators’ viewing, reaching, and sample-loading actions. This study examined a concept-stage workflow that combined a low-fidelity bench with single-smartphone markerless motion capture before a complete engineering prototype was available. Interviews with seven engineers and task analysis were used to define posture-risk criteria, which were prioritized using the analytic hierarchy process (AHP). Twenty-five design students served as proxy operators. OpenCap-derived trajectories were processed in Python to calculate sagittal-plane joint angles and Rapid Upper Limb Assessment (RULA)-oriented screening indicators. In the primary near-sagittal subgroup (β = 0°, n = 14), the mean raw 99th percentile of neck flexion was 52.53° ± 7.55°, and all 14 participants exceeded the predefined 45° RULA-oriented threshold. The primary AHP–RULA analysis was restricted to this subgroup and provided a directional internal cross-check; the full-sample analysis was retained only as a sensitivity analysis. The workflow may support early layout screening and formative-evaluation documentation. Because no concurrent reference measurement system was used, the angle values should be interpreted as screening estimates rather than validated absolute measurements. The workflow does not replace usability validation conducted with intended clinical users and an engineering prototype.