Dr Oren Griffiths

Personnel are faced with an ever-increasing amount of information to process. Highly demanding tasks, such as monitoring for threats, operating complex machinery and system maintenance put significant load on the brain’s ability to attend to what is important. To reduce the attentional burden and increase performance, Augmented Reality (AR) solutions are being developed. This technology superimposes a computer-generated image on a user's view of the world, thus providing a composite view. Such information is rapidly being incorporated into many complex task environments to provide “heads-up” information where and when it is needed. The promise of AR is enormous.

Despite the promise of AR, research shows that it can impair performance, rather than improve it. For example, in a study using military personnel, the reliability of cues to locate camouflaged targets was manipulated. Results showed that cuing improved detection for expected (cued) events – but led to misses on unexpected events. Similarly, onscreen overlays have led pilots to fail to detect obstacles while landing, and surgeons to miss critical complications. The common link in these situations is that AR cues can capture the user’s attention so effectively that unanticipated, but critical, events are missed. This phenomenon is referred to as “inattentional blindness” and is thought to underlie many civilian motor accidents in which motorists collide with a plainly visible object. Despite the importance of these “looked but didn’t see” events, the phenomenon can be difficult to study in applied contexts (e.g. in cockpits), because the standard procedure to measure attention is eye-gaze. Inattentional blindness is invisible to eye-gaze because those who miss the target are just as likely to gaze at it as those who notice it.

Our research is directly relevant to the use of technologies to enhance human performance. Specifically, we will provide a better understanding of “looked but didn’t see” errors caused by AR cueing by employing a novel brain-based measure of attention, which can operates independently of gaze (covert attention). Covert attention can be measured using steady state visual evoked potentials (SSVEP) on an electroencephalograph (EEG). This method exploits a quirk of the neural visual processing system. When a person attends to a flickering stimulus, the frequency of that flickering stimulus is heightened in resulting neural activity. We will use this technique along with quantitative and qualitative measures to assess human performance in AR.