Ruth Aylett

BACKGROUND Virtual reality (VR) technologies offer opportunities to assess and support activities of daily living in ecologically valid, controllable, and adaptable environments. However, many available systems remain theory-light, insufficiently personalised, and limited in their capacity to map cognitive processes onto everyday functional performance across age. OBJECTIVE This study aimed to develop and evaluate a cognitive-oriented framework for assessing the cognitive underpinnings of everyday kitchen-based activities in real and non-immersive VR environments, and to determine how performance varies by age and environment. METHODS A between-subjects experimental design was used. One hundred healthy participants were randomly assigned by age group and environment to four groups: older real environment (ORE; n=28; mean age 77, SD 4.75), younger real environment (YRE; n=24; mean age 27.61, SD 5.47), older VR environment (OVE; n=24; mean age 74.75, SD 5.33), and younger VR environment (YVE; n=24; mean age 21.29, SD 2.68). Participants completed a kitchen-based Multiple Errands Task comprising 6 errands and 54 activities in either a real kitchen or an equivalent non-immersive VR kitchen along with a brief neuropsychological assessment for older adults. Performance was scored for correct actions, completion time, and error types, including misses, order, executive, semantic, episodic, self-corrected, errand-order, and intrusion errors. General linear models tested effects of age group, environment, and their interaction; Pearson correlations, corrected using false discovery rate procedures, examined associations between neuropsychological measures and task performance. RESULTS Older adults completed fewer correct actions and made more total errors, misses, order errors, intrusions, and errand-order errors than younger adults. VR increased task demands: both age groups took longer in VR than in the real environment, and older adults were particularly slower in VR, yielding a significant Group × Environment interaction. Correctly performed actions were lower in older than younger adults, with older adults performing significantly better in the real environment than in VR, whereas younger adults’ accuracy was not significantly affected by environment. Interference did not significantly affect accuracy. Processing speed was positively associated with correct performance and negatively associated with total errors in the real environment; in older adults in VR, Mini-Mental State Examination scores correlated negatively with self-corrected errors (r=-.808, P<.001, n=24), and study time correlated positively with episodic errors (r=.641, P=.001, n=24). Older adults also made more task-related intrusions than younger adults across environments. CONCLUSIONS The VRAIS framework identified age-sensitive cognitive-functional performance patterns in real and VR contexts. Findings suggest that VR can reveal meaningful cognitive demands of everyday tasks, but ecological validity depends not only on environmental realism, but also on personal familiarity, lifelong behavioural schemas, and preparation. Theory-driven, adaptive, and person-centred VR systems may support more precise assessment and future cognitive rehabilitation for ageing and dementia-related functional decline.

The AMPER application is designed to support reminiscence therapy in a domestic setting for those living with dementia through the use of an Intelligent Virtual Agent (IVA). Such agents have been shown to increase functionality, accessibility and user satisfaction. We describe the application and summarise the design process used to choose the graphical agent for the project. We briefly discuss the implications of using stereotypes or archetypes in this process.

This paper reports work addressing the research question "could a Bechdel-like test be an indication of how women are represented in video games?" through developing such a test for video games: the Indicative Representation of Women In Games (IRWiG) test. We describe its development process: a multidisciplinary approach combining a literature review with the development and application of ontologies representing the constructional elements of games that relate to portrayal of women. The IRWiG test was evaluated through a public evaluation survey and an internally-conducted games analysis. The test proposes four criteria for analysing a female character: her character development, appearance, abilities and relevant stereotypes, and the skippability of content in which she is active. An overall agreement rate of 74% was found between users' opinions of how a woman is represented in a game, and the application of the IRWiG test.

We present a demonstration of the AMPER system: an Android application that guides individuals with Alzheimer’s disease and their carers through reminiscence therapy with the use of a virtual agent. The application supports a novel method of reminiscence story selection, using material metadata, user data, and a spreading activation algorithm. This aims to present relevant material to the user, in an order that attempts to mimic a human autobiographical memory.

The present study describes the process employed to identify a suitable intelligent virtual agent (IVA) for the AMPER App supporting reminiscence therapy for those living with dementia through the use of a facilitating agent. This included three distinct phases: 1) co-creation with project stakeholders and identification of a set of desirable IVA traits; 2) a blind internal team rating process to select a subset from available IVAs; 3) a character survey with healthy older adults to select a final 4 IVAs (2 male, 2 female). We analyse the results, assess inter-subject agreement, and suggest guidelines for IVA selection.

Experiential artificial intelligence (AI) is an approach to the design, use, and evaluation of AI in cultural or other real-world settings that foregrounds human experience and context. It combines arts and engineering to support rich and intuitive modes of model interpretation and interaction, making AI tangible and explicit. The ambition is to enable significant cultural works and make AI systems more understandable to nonexperts, thereby strengthening the basis for responsible deployment. This paper discusses limitations and promising directions in explainable AI, contributions the arts offer to enhance and go beyond explainability and methodology to support, deepen, and extend those contributions.

Experiential AI is an emerging research field that addresses the challenge of making AI tangible and explicit, both to fuel cultural experiences for audiences, and to make AI systems more accessible to human understanding. The central theme is how artists, scientists and other interdisciplinary actors can come together to understand and communicate the functionality of AI, ML and intelligent robots, their limitations, and consequences, through informative and compelling experiences. It provides an approach and methodology for the arts and tangible experiences to mediate between impenetrable computer code and human understanding, making not just AI systems but also their values and implications more transparent, and therefore accountable. In this paper, we report on an empirical case study of an experiential AI system designed for creative data exploration of a user-defined dimension, to enable creators to gain more creative control over the AI process. We discuss how experiential AI can increase legibility and agency for artists, and how the arts can provide creative strategies and methods which can add to the toolbox for human-centred XAI.

We focus on affective architecture issues relating to the generation of expressive facial behaviour, critique approaches that treat expressive behaviour as only a mirror of internal state rather than as also a social signal and discuss the advantages of combining the two approaches. Using the FAtiMA architecture, we analyse the requirements for generating expressive behavior as social signals at both reactive and cognitive levels. We discuss how facial expressions can be generated in a dynamic fashion. We propose generic architectural mechanisms to meet these requirements based on an explicit mind-body loop and Theory of Mind (ToM) processing. A illustrative scenario is given.