Original title: Evaluation of visual guidance techniques for augmented reality-supported assembly tasks in the medical context
Julian Kreimeier, Prathik Prasad, Shiyu Li, Niklas Corell +8 · September 2026
Putting surgical tools together correctly is vital for patient safety and smooth medical workflows. However, assembling them by hand can be difficult. Researchers tested whether augmented reality (AR)—which overlays digital graphics onto the real world—could guide people through assembly tasks better than traditional paper manuals. The system tracks parts without special markers, shows step-by-step visual guidance, and lets users advance through steps using their eye gaze. The researchers compared three visual display styles: placing digital animations directly over the physical parts, fixing them in the user's field of view, or anchoring them in the surrounding room. They found that AR guidance matched and partly outperformed paper instructions in workload, usability, and user experience. Times and errors were largely comparable to paper. Anchoring animations in the room or user's field of view worked better than placing them directly on parts, which suffered from tracking inaccuracies.
You are assembling a piece of flat-pack furniture. The paper instructions show each step, but you have to keep looking away from the parts to check the diagram, and sometimes the perspective is different from what you see. In this paper, the instructions are shown directly on the parts through a headset, so you always see exactly where each piece goes from your own viewpoint, without having to interpret a 2D drawing.
For scrub nurses assembling surgical instruments, paper instructions are standard but can be slow and error-prone. This study shows that augmented reality guidance—using head-mounted displays—can match or beat paper in speed and accuracy. That means safer surgeries and faster training, especially for complex or rarely used instruments.
The study used 3D-printed proxies, not actual surgical instruments; the sample size is 32; the task was a benchmark, not a real clinical workflow; in-situ animations suffered from pose estimation inaccuracies; objective metrics were largely comparable, not superior; and individual differences suggest a need for personalization.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Evaluation of visual guidance techniques for augmented reality-supported assembly tasks in the medical context — Julian Kreimeier, Prathik Prasad, Shiyu Li et al. — Virtual Reality — doi:10.1007/s10055-026-01491-3
Abstract licence: CC BY
Explanation generated by Synthesis from the abstract — it is not the original paper.