AI-Supported Question-Based Learning (pQBL) Modules for Rehabilitation and Assistive Technologies
A technology-enhanced educational deliverable developed by KTH Royal Institute of Technology as part of the European Dual Master Programme in Rehabilitation and Assistive Technologies — in collaboration with the University of Macedonia (UoM) and the Aristotle University of Thessaloniki (AUTH).
Developed by Prof. Olof Bälter, Professor in Computer Science at KTH Royal Institute of Technology, Media Technology & Interaction Design.
KTH Royal Institute of Technology
Dual Master Programme
EU Collaboration
About the Developer
KTH Royal Institute of Technology
This deliverable was developed by Prof. Olof Bälter, Professor in Computer Science at KTH Royal Institute of Technology (Media Technology & Interaction Design). Olof Bälter is the founder of the KTH research group for Technology Enhanced Learning and the creator of the Pure Question-Based Learning (pQBL) methodology.
Overview
This deliverable presents the development of a technology-enhanced pedagogical package grounded in the pure Question-Based Learning (pQBL) approach, augmented by AI-assisted question generation. The goal is to transform existing course content into structured, interactive learning units that actively engage students through guided questioning, immediate feedback, and progressive skill development. The deliverable includes fully developed pQBL modules already deployed on the Torus platform, along with structured skillmaps and AI-ready input files that enable immediate use and further content generation.
Directly Usable
Ready for deployment in Master-level teaching from day one, with no additional adaptation required.
Institutionally Adaptable
Designed to be reused and adapted across all three partner institutions seamlessly.
Field-Aligned
Grounded in real-world applications in rehabilitation and assistive technologies.
Pedagogical Approach: pQBL
The pure Question-Based Learning (pQBL) methodology places structured questioning at the center of the learning experience. Rather than passively consuming content, students actively make decisions, confront misconceptions, and build understanding step by step.
Core Principles
  • Structured sequences of questions driving exploration
  • Immediate, targeted feedback per response
  • Progressive development of understanding and applied skills
Especially Suited For
  • Complex, applied academic domains
  • Interdisciplinary learning environments
  • Heterogeneous student backgrounds and prior knowledge
Technological Component
The deliverable integrates cutting-edge AI tools with a dedicated web-based platform to enable scalable, efficient content development. The Torus/pQBL platform serves as the backbone for generating, editing, and deploying all learning modules.
AI-Assisted Question Generation
Questions are generated automatically from learning objectives and structured course material, dramatically reducing content development time.
Torus/pQBL Platform
A web-based system enabling instructors to generate, review, edit, and deploy question-based modules with minimal technical overhead.
Scalability & Reuse
Content can be rapidly developed and reused across multiple courses and partner institutions, maximizing return on investment.
Connection to Course Material: Research Methodology & Biostatistics
Primary Source Material
The pQBL modules are directly grounded in the existing curriculum of the Master programme. The most comprehensive source material comes from the Research Methodology and Biostatistics course, which provided lecture notes, weekly worksheets, practical guides, and seminar content — all transformed into structured learning units.
Material Sources
  • Lecture notes and slides
  • Weekly worksheets and exercises
  • Practical guides and seminar material
Key Topics Covered
  • Research design and methodology
  • Sampling in vulnerable populations
  • Measurement validity and reliability
  • Statistical testing: ANOVA, regression, and longitudinal analysis
  • Interpretation of clinical outcomes and results
From Course Content to pQBL Skillmap
These topics are directly reflected in the generated pQBL skillmap and learning units, which emphasize applied decision-making and analysis through realistic research scenarios. Rather than focusing only on theory, the modules guide learners through authentic problems that require selecting methods, interpreting evidence, and evaluating clinical findings in context.
Research Methodology & Biostatistics — Course on OLI Torus
Live on OLI Torus Platform
The course is fully deployed on the OLI Torus platform, where students access structured pQBL units progressively. The screenshot below shows the course dashboard as experienced by students.
01
Intro to pure question-based learning (pQBL)
(Practice)
02
Introduction to the scientific method and research question formulation (PICO)
(Lesson)
03
Research designs in rehabilitation
(Lesson)
04
Sampling, ethics, and data management
(Lesson)
05
Clinical scales, questionnaires, and psychometric properties
(Lesson)
06
Descriptive statistics and data visualization
(Lesson)
Connection to Course Material: Assistive Technologies & Accessibility
Course Description-Based
Based on the official Assistive Technologies and Accessibility course descriptions, this module stream defines conceptual learning objectives and establishes the domain context for AI-driven question generation. The focus is on building students' ability to critically analyze accessibility challenges, user needs in everyday environments, and propose appropriate solutions in daily living contexts.
Accessibility Barriers
Identifying and classifying physical, cognitive, and digital barriers faced by individuals with disabilities.
Assistive Solutions
Understanding assistive devices, software tools, and environmental modifications that promote independence.
Universal Design Principles
Applying design-for-all frameworks to create inclusive products, services, and environments.
Connection to Course Material: Rehabilitation Engineering
Applied & Technological Perspectives
The Rehabilitation Engineering course descriptions provided the applied technological lens for this deliverable. This stream extends student understanding into emerging, data-driven paradigms, preparing graduates to operate at the intersection of engineering, AI, and clinical rehabilitation practice.
AI in Rehabilitation
Exploring how machine learning and AI tools are transforming assessment, therapy, and outcome monitoring in rehabilitation contexts, including adaptive decision support and automated progress tracking.
Data-Driven Personalization
Understanding how patient data enables the customization of rehabilitation pathways and therapeutic interventions, informed by sensors, performance metrics, and longitudinal feedback.
Intelligent Assistive Systems
Examining smart, responsive assistive technologies that adapt to individual user needs in real time, including robotics, sensor-based devices, and brain-computer interfaces.
Connection to Course Material: Introduction to Disability & Rehabilitation
Conceptual & Social Foundations
The Introduction to Disability and Rehabilitation course enriches the pQBL modules by incorporating perspectives on disability models and participation barriers. This course provides the conceptual and social foundation that contextualizes all applied learning units within the programme.
Disability Models
Exploring medical, social, and biopsychosocial models of disability and their implications for rehabilitation practice and policy.
Participation Barriers
Analyzing environmental, attitudinal, and systemic barriers that limit the full participation of individuals with disabilities in society.
Rehabilitation Frameworks
Understanding the goals, principles, and interdisciplinary approaches that guide contemporary rehabilitation practice.
These perspectives ensure that pQBL units go beyond technical knowledge, fostering critical awareness of the social and human dimensions of rehabilitation.
Connection to Course Material: Multimodal Interaction & Interfaces
Interaction Design & Sensory Engagement
The Multimodal Interaction and Interfaces course further enriches the pQBL context by incorporating multisensory interaction design principles. This course bridges human-computer interaction with rehabilitation, ensuring that learning units address how users with diverse abilities engage with technology.
Multisensory Interaction Design
Understanding how visual, auditory, tactile, and gestural modalities can be combined to create accessible and effective user interfaces.
Inclusive Interface Development
Applying interaction design principles to develop interfaces that accommodate users with varying sensory, motor, and cognitive abilities.
Human-Technology Integration
Examining how multimodal systems support rehabilitation goals by enabling more natural and adaptive forms of human-computer interaction.
These insights are integrated into pQBL units that address interface design challenges in assistive and rehabilitation technology contexts.
Structure of the Learning Modules
Each topic is encapsulated in a pQBL learning unit built around a consistent, pedagogically sound structure. Every unit is carefully aligned with specific learning goals and designed to promote deep conceptual understanding rather than surface-level recall.
Unit Architecture
  • 4–6 questions per learning unit
  • Progressive difficulty from foundational to applied
  • Direct alignment with specific course learning goals
Question Design
  • 3 answer alternatives per question
  • Plausible distractors based on common misconceptions
  • Targeted, explanatory feedback for each option
  • Learning rationale embedded within feedback text
This carefully crafted structure ensures that every student interaction is a meaningful learning event, not merely a test of recall.
Deliverable Components
The deliverable is a fully developed, ready-to-use educational package — not a prototype or framework. It includes the following five components, all of which are already operational on the Torus platform:
1
Structured pQBL Skillmap
Learning objectives and domain-specific skills mapped directly to course material across all three subject areas.
2
AI-Ready Input Material
Structured texts, defined learning objectives, and module specifications formatted for AI-assisted question generation.
3
Generated Learning Units
Complete question-based units, ready to use in live teaching and aligned with programme course content.
4
Platform Integration
Full compatibility with the Torus/pQBL system, with potential for broader LMS integration (e.g., Moodle, Canvas).
5
Documentation
Usage guidelines, instructional support materials, and concrete implementation examples for educators.
Use within the Master Programme
The deliverable is designed with flexibility as a core feature. It can be deployed in multiple pedagogical modes within the Master in Rehabilitation and Assistive Technologies programme, supporting both structured instruction and independent student learning.
Interactive Teaching Modules
Used live in class as the primary instructional tool, replacing or complementing traditional lectures.
Self-Paced Learning
Students engage with modules independently at their own pace, reinforcing understanding outside the classroom.
Pre- & Post-Lecture Reinforcement
Activates prior knowledge before class or consolidates learning immediately after a lecture session.
Formative Assessment
Provides instructors with a structured, low-stakes tool for evaluating student understanding and identifying gaps.
Added Value of This Deliverable
Beyond its immediate educational utility, this deliverable represents a significant methodological contribution to the programme and to the broader European consortium. It demonstrates how technology and pedagogy can be combined purposefully to create inclusive, scalable learning experiences.
A Scalable Transformation Method
Provides a replicable workflow for converting any course content into engaging, question-based learning experiences — usable by all partner institutions.
A Bridge Between Pedagogy and Technology
Demonstrates how AI tools and digital platforms can serve genuine educational goals rather than technology for its own sake.
A Reusable Framework
The structure, skillmaps, and documentation can be adapted to new courses at KTH, UoM, AUTH, or any future programme partner.

Most importantly, this deliverable enables: active, structured, and inclusive learning in a complex and rapidly evolving field — placing students at the center of their own educational journey.
Future Development
This deliverable establishes a strong and extensible foundation. The following directions represent natural next steps for expanding impact within the programme and beyond, ensuring the pQBL framework continues to evolve alongside the field of rehabilitation and assistive technologies.
1
Course Expansion
Extend pQBL modules to additional courses across all three partner institutions.
2
LMS Integration
Full integration with platforms such as Moodle and Canvas for seamless institutional deployment.
3
Adaptive Learning Pathways
Develop personalized question sequences that adapt dynamically to individual student performance and needs.
4
Student Analytics
Implement performance tracking and learning analytics dashboards to inform teaching and improve outcomes.
Developed by KTH Royal Institute of Technology in the framework of the Dual Master in Rehabilitation and Assistive Technologies, in partnership with the University of Macedonia (UoM) and the Aristotle University of Thessaloniki (AUTH).
Reference Material
Documentation & Technical Guide
Full technical and educational documentation for the AI-Supported pQBL Modules in Rehabilitation and Assistive Technologies. Select a section below to expand.
Platform Overview (OLI Torus)
The pQBL modules are hosted on the OLI Torus platform, an open-source adaptive learning environment developed by Carnegie Mellon University. Torus supports question-based learning units, skill mapping, and detailed student analytics.
AI-Assisted Question Generation Process
Questions were generated using large language model (LLM) tools, guided by structured prompts derived from course material. Each generated question was reviewed, refined, and aligned with specific learning objectives before deployment.
Skillmap Structure
Each course is mapped to a set of learning objectives (skills). The skillmap defines the relationship between questions and skills, enabling adaptive sequencing and targeted feedback within the platform.
Unit Architecture
Each pQBL unit contains 4–6 questions with 3 answer alternatives each. Questions progress from foundational to applied difficulty. Each answer option includes targeted explanatory feedback and an embedded learning rationale.
Course Units — Research Methodology & Biostatistics
  • Unit 1 (Practice): Intro to pure question-based learning (pQBL)
  • Unit 2 (Lesson): Introduction to the scientific method and research question formulation (PICO)
  • Unit 3 (Lesson): Research designs in rehabilitation
  • Unit 4 (Lesson): Sampling, ethics, and data management
  • Unit 5 (Lesson): Clinical scales, questionnaires, and psychometric properties
  • Unit 6 (Lesson): Descriptive statistics and data visualization
Partner Institutions
  • KTH Royal Institute of Technology (Sweden) — lead developer
  • University of Macedonia (UoM) — Greece
  • Aristotle University of Thessaloniki (AUTH) — Greece
Intended Use & Deployment
The modules are designed for Master-level students in the Dual Master Programme in Rehabilitation and Assistive Technologies. They can be deployed directly via OLI Torus or adapted for integration with Moodle or Canvas.
Licensing & Reuse
All materials are developed within the framework of the European Dual Master Programme and are intended for open academic reuse across partner institutions.