Despite advances in educational technology, university teaching remains largely designed for large cohorts and standardised delivery, overlooking the diverse learning needs of individual students. Learners with ADHD or other special educational needs often struggle to engage effectively in such rigid and non-personalised environments.

This raises a critical question: can we design immersive classrooms that adapt to individual cognitive needs?

The COVID-19 pandemic forced universities to rapidly adopt remote teaching at scale, exposing several limitations—impersonal delivery, dependence on devices and internet connectivity, and a lack of real-time personalisation. Additionally, online platforms are poorly suited to demonstrating in-class experiments or practical exercises, which are essential for experiential and active learning.

Current educational content is still predominantly delivered through text and video on flat screens. While this approach is convenient, it fails to support the understanding of complex 3D concepts, particularly in subjects such as anatomy, architecture, or engineering, where spatial reasoning is vital. Learners often struggle to grasp depth, scale, and the relationships between objects when content is limited to two-dimensional formats (Huang & Tseng, 2025).

Immersive learning environments offer a promising solution by providing interactive and spatially rich experiences. However, these systems are often too complex and lack personalised guidance and  real-time feedback. This leads to disengagement and suboptimal learning outcomes. To address these challenges, conversational AI and neurofeedback can be integrated to enhance learning, enabling real-time adaptation to the learner’s cognitive and emotional state (González-Erena, Fernández-Guinea, & Kourtesis, 2024).

This PhD research proposes an adaptive learning framework within Mixed Reality (MR) environments, integrating an AI-enhanced virtual tutor and an exploratory neurofeedback component. The neurofeedback, collected via a brain–computer interface (BCI), is intended as an investigation into whether real-time neurological data can be used to inform and enhance the design of learning systems in immersive environments (see Figure 1). Together, these technologies are capable of creating an interactive one-to-one learning experience tailored to the learner’s individual profile. By continuously monitoring the user’s conversation with the AI tutor, alongside cognitive engagement through neurofeedback, the system can provide adaptive support to learners, with the aim of improving engagement, retention, and overall learning outcomes in immersive educational settings.

To facilitate this investigation, several immersive environments will serve as research instruments. They will provide crucial insights into how the integration of these technologies can shape the future of personalised education, exploring their potential for real-world applications and demonstrating their effectiveness.