Trainees help design the tool, which tells us about usefulness but not about learning
In surgical training, the case for AR eye-gaze tracking has been built by working directly with the trainees who would use it. A co-design study with surgical trainees produced user-informed guidelines for future collaborative AR-based eye-gaze tracking in endoscopic instruction, aimed at improving training modules and integrating AR into operating-room teaching [1]. That is evidence about what trainees and designers think the system should do — their needs and priorities — rather than a test of whether using it makes anyone more skilled.
The same pattern appears in other training domains. A tram driver simulator project combined simulator technology, augmented reality, and eye tracking to target visual attention, addressing what the authors describe as a recognized need in transportation training [2]. The concept and early feedback are reported, but the work does not include quantitative measures of learning gains or training effectiveness [2]. So the enthusiasm is real and grounded in a genuine training problem; it just hasn't been cashed out as a measured improvement.
Why gaze might be a useful training signal in the first place
The underlying idea is that where someone looks reveals something about how they are paying attention, and attention is part of what training is trying to shape. In extended reality, eye-tracking data has been used to study attention patterns in AR and VR, and machine-learning models trained on that data could predict whether distractors were present and how well users controlled their attention [5]. If gaze can flag when a trainee is being pulled off task, an instructor or a system could in principle respond in the moment.
That is a plausible route from tracking to teaching, but it is a possibility rather than a demonstrated benefit. The attention work shows that gaze carries information about attention in these environments [5]; it does not show that feeding that information back to a trainee improves their performance. Reviews of eye-tracking in AR and VR note its role in aviation and military training and the real-time gaze tracking that interactive AR depends on [4], which explains why the approach keeps attracting investment even while outcome data lag behind.
The engineering is further along than the teaching evidence
A good part of the published work is about making the hardware work at all. One surgical training platform paired the HoloLens augmented reality headset with a separate eye-tracker, with attention to system design and calibration [3]. Calibration matters because gaze tracking is only useful if the system knows reliably where the trainee is looking; that is an engineering problem, not a pedagogical one.
This is the gap that makes the question hard to answer cleanly. The design studies tell us the tool can be built and that trainees have opinions about how it should work [1][3]; the simulator work tells us a combined system can be implemented and demonstrated [2]; the review work tells us the technology has a foothold in high-stakes training [4]. What none of it supplies is the comparison a curious reader would most want: trainees trained with AR eye-gaze feedback versus trainees trained without it, measured on the same skill.
What would actually settle whether it helps people learn
The most direct evidence in this set is about attention, and it points to a way forward rather than a verdict. Because gaze-based models can predict distraction and attention control in AR and VR [5], a future study could use those predictions as the intermediate measure and then check whether trainees whose attention is flagged end up performing better. Until that link is tested, attention remains an indicator that researchers treat as relevant to training effectiveness, not proof that training improved.
For a reader deciding whether to trust the pitch, the practical reading is this: the usefulness claims rest on trainee co-design and early feedback [1][2], and the learning claims rest on attention as a stand-in for skill [5]. Both are reasonable starting points, and neither is the same as showing gains. The honest answer to whether trainee enthusiasm matches actual learning is that the enthusiasm is documented and the learning is not yet measured in these studies.
Sources used in this answer
Focus on the experts: Co-designing an augmented reality eye-gaze tracking system with surgical trainees to improve endoscopic instruction
A co-design study with surgical trainees generated user-informed guidelines for designing future collaborative AR-based eye-gaze tracking systems for endoscopic instruction, focusing on user needs rather than learning outcomes.
Augmented Reality Eye Tracking for Tram Driver Simulator Training: Concept, Implementation, and Early Feedback
A tram driver simulator training system combining simulator technology, augmented reality, and eye tracking was introduced with its concept, implementation, and early feedback, but without quantitative measures of learning gains or training effectiveness.
Integrating eye-tracking to augmented reality system for surgical training
A surgical training platform combined the HoloLens augmented reality system with an eye-tracker, emphasizing system design and calibration rather than empirical measurements of learning gains.
Eye-tracking in ar/vr: A technological review and future directions
A technological review reports that eye-tracking in AR/VR contributes to training in aviation and military contexts and that real-time gaze tracking is important for interactive AR, without presenting empirical data on learning gains.
“Looking” into attention patterns in extended reality: An eye tracking-based study
Machine-learning models trained on eye-tracking data from AR and VR environments could predict the presence of distractors and users' attention control ability, with performance dropping when transferred between conditions.