A methodology for assessment of spatial performance in hospital buildings through immersive virtual reality and behavioural sequence analysis

Abstract

The spatial and wayfinding performance of a building are key elements that affect how well a building functions. They play a crucial role in buildings' overall functionality, particularly in complex buildings like hospitals that contain interconnected units. Time loss and detrimental impact on user well-being arise when a building's wayfinding performance is insufficient. Understanding and incorporating wayfinding behaviour into the design process is essential for designing human-centred buildings. This study examines wayfinding behaviour in detail to comprehend the dynamics behind wayfinding thinking and its relationship with spatial characteristics. The thesis questions (1) the effects of spatial plan configuration on wayfinding behaviour, (2) the relationship between re-occurring behaviours during the wayfinding process depending on spatial characteristics, (3) the extent and level of detail in behavioural data that can be gathered about human wayfinding through virtual reality (VR). For this purpose, this study utilises Behavioural Sequence Analysis (BSA) and Space Syntax analysis to understand the relationship between wayfinding behaviours and spatial layout in hospital buildings through Immersive Virtual Reality (iVR). BSA investigates behavioural sequences and identifies behavioural transition patterns. Space syntax analysis measures spatial features quantitatively. This study links wayfinding behaviours with spatial features and each other by employing these two analyses. An experiment was designed employing iVR since it delivers environments that are remarkably close to reality and brings together conditions that cannot be brought together in real life. The experiment aimed to observe and record user behaviour during the wayfinding process. Hospital buildings with three different plan layouts based on the centrality of the spatial configuration were selected and space syntax analysis was used to create visibility maps of these buildings, which were subsequently utilised for final assessments. Virtual environments (VE) were prepared by creating partial 3D models of the buildings with the architectural information collected for the experiment. Wayfinding tasks with various target locations were given to the participants in each hospital building during the iVR experiment to gather behavioural data. The experiment was recorded and the routes of the participants were mapped on floor plans. After analysing experimental records, the relevant data was extracted and arranged for four types of analyses: (1) BSA, (2) visibility and behavioural relationship analysis, (3) behavioural frequency analysis and (4) average speed analysis. These four analyses investigated how wayfinding behaviours varied by space, how often they occurred, which trends emerged in behavioural sequence transitions, and how quickly participants moved through each building. Following an evaluation of each analysis, a holistic approach was applied to these findings questioning the reliability of each analysis. The findings showed that each type of plan layout directs the users to engage in different behavioural groups during the wayfinding process. The recurring transitions among the behaviours revealed the wayfinding thinking of participants for each hospital building. The centralised plan layout directs the users to a fast-paced decision-making process, whereas the semi-centralised leads to a slow and efficient one. On the other hand, the findings demonstrate that there is a directly proportional relationship between visibility and wayfinding performance. Integrated analysis results show that the semi-centralised plan layout with high visibility offers the most comfortable and efficient wayfinding experience comparatively and the decentralised plan layout with low visibility offers the worst. The results can be used to create human-centred buildings by employing the spatial characteristics that provide the most effective wayfinding outcomes, particularly in the early design stages. Furthermore, the spatial-behavioural connections detailed in this research can be used as a guide depending on the building typology and the desired user experience.