Thesis defence : Sarah KHAZAZ

Recent studies have shown that in stressful, high-risk situations, building emergency evacuation systems, particularly emergency lighting, are not always easily perceived and followed by evacuees. Yet the human stakes involved in a well-guided and successful evacuation are considerable. At present, this work (1) does not consider the visual and attentional system of human visual perception (2) essentially tests evacuation scores and self-reported questionnaires (3) does not measure or model the effect of psychological stress linked to evacuation on detection, comprehension and application of an evacuation instruction issued by a sign or emergency lighting fixture (4) do not explain the neurocognitive mechanisms involved in the detection of an emergency evacuation sign and do not quantify the effect of the characteristics of the evacuation sign on the cerebral processing of the information. This CIFRE thesis project is part of a market-leading emergency lighting manufacturer's desire to pursue research aimed at improving the detectability of evacuation luminaires with a view to increasing their effectiveness. The aim of the work is to better understand and specify the psychophysical, behavioural and neurophysiological mechanisms that determine the detection, understanding and application of an evacuation instruction provided by an emergency lighting luminaire.

In the first part of the thesis, in the light of the existing literature on the various characteristics relevant to the detection of an emergency lighting system, we asked in a first experiment whether the variables of target shape and colour could influence rapid detection in peripheral vision on the one hand and search in central vision on the other. Similarly, in a second experiment, we wanted to assess the extent to which the contrast and flashing of a target influence their detection. And finally, in a third experiment, we tried out new types of pictograms (aversive and humanoid). Our aim was to make maximum use of an environment representing an evacuation zone. To do this, we used artificial visual scenes, in other words images of stations, supermarkets or airports. We chose complex visual scenes, in other words scenes that contained a lot of distracting information in terms of colour and shape.

In the second experimental part of the thesis, the aim was to investigate psychological stress more fully. Indeed, we felt that it was relevant to study the involvement of stress in relation to attention, as stress could have a significant impact on the way in which we perceive, process and control our attention. It therefore seemed crucial to us to assess the extent to which the psychological stress to which an individual is subjected during an emergency evacuation situation impacts on perception and attention. More specifically, we wanted to test the influence of stress on the detection of the same psychophysical characteristics that we evaluated in our experiments (shape, colour, contrast and temporal frequency). To induce stress during this protocol, we interrupted the trials by the unpredictable occurrence of the simultaneous appearance of frightening visual and auditory stimuli (< 80 dB) for two seconds. In order to quantify the changes in participants' performance under psychological stress, we conducted two experimental conditions: a control condition (without the appearance of frightening stimuli) followed by a stress condition (with the appearance of frightening stimuli).

In a third experimental part, we wanted to extend our behavioural results to an emergency evacuation situation. This part was a major focus of the thesis, as we were asked to apply all our results in a project to create a virtual environment where we could test new emergency lighting prototypes inspired by our behavioural results. We mainly highlighted our results on the detection of a flashing light at 10 Hz, which we judged to be optimal in a situation without stress and in a situation with psychological stress.

Overall, this thesis has extended the field of emergency lighting research by incorporating results from the methodology applied in cognitive psychology, which plays a fundamental role in the way we design and optimise lighting to better meet human needs. This connection between cognition and human-centred lighting allows us to improve our interaction with the visual environment by taking advantage of scientific knowledge to create more functional signage.

- Directeur de thèse : Martial MERMILLOD - martial.mermilloduniv-grenoble-alpes.fr (martial[dot]mermillod[at]univ-grenoble-alpes[dot]fr)
- Codirecteur : Sébastien POINT - 0381666738 - sebastien.pointcnrs.fr (sebastien[dot]point[at]cnrs[dot]fr)
- Co-encadrant : LAURIE MONDILLON - 473406107 - Laurie.MONDILLONuca.fr (Laurie[dot]MONDILLON[at]uca[dot]fr)
- Co-encadrant : Pierre-Olivier DEFAY - Pierre-OlivierDefayEaton.com (Pierre-OlivierDefay[at]Eaton[dot]com)