Browsing by Person "Shidende, Deogratias"

Now showing 1 - 1 of 1

Design and development of an accessible open-source augmented reality learning authoring tool for applications in agroecological settings
(2024) Shidende, Deogratias; Treydte, Anna
Augmented Reality (AR) has emerged as a transformative educational technology, offering immersive, multisensory learning experiences that enhance engagement, conceptual understanding, and contextualization. In agroecology, where students must grasp complex ecological interactions and context-dependent knowledge, AR can bridge the gap between classroom instruction and field-based learning. However, the creation of AR content remains largely inaccessible to many educators in higher learning institutions (HLIs), particularly those without programming skills and individuals with disabilities such as the deaf and hard of hearing (DHH), and the blind and low vision (BLV). This dissertation addresses the central question: How can an accessible AR learning authoring tool enable non-technical educators and users with disabilities to create AR learning experiences for agroecology education in HLIs? To address this question, the study employed a design-based research (DBR) methodology, integrating Agile Scrum for iterative, inclusive tool development. Seven research questions (RQ1–RQ7) guided the investigation. First, a document-based analysis (RQ1) compared open-source software licenses (OSLs) to determine their suitability for academic–industry collaboration. Permissive licenses (e.g., MIT, BSD) were found to offer more flexibility in code reuse and integration, thereby promoting long-term project sustainability, although they require supplementary legal mechanisms to ensure reciprocity. Next, functional and non-functional software requirements (RQ2) were elicited through stakeholder workshops, interviews, surveys, and accessibility evaluations. These requirements informed the selection and redesign of MirageXR, an open-source AR platform. Key accessibility features were specified for DHH users, such as customizable captioning of audio augmentations, and for BLV users, such as voice navigation and spatial audio cues. These enhancements underscored the dual instructional and assistive roles of AR tools. In response to RQ3, a modular, component-based software architecture was designed using the C4 model. This enabled seamless integration of external services (e.g., 3D object repositories, learning management systems, and automatic speech recognition) and ensured that features could be added or updated without disrupting system stability. This modularity was essential given the evolving nature of AR technologies. The design and implementation phases (RQ4 & RQ5) employed participatory iterative prototyping with user feedback throughout the development process. Accessibility features were integrated into image, audio, and video augmentations, with functionalities such as caption editing, playback speed control, and 3D spatial positioning. These solutions directly addressed gaps in existing AR authoring tools, particularly for DHH and BLV users. The sixth research question (RQ6) investigated usability and applicability through an AR creation workshop involving 24 agroecology educators. Findings revealed that although participants initially encountered difficulties, they gained proficiency over time. UMUX scores showed a correlation between AR experience and perceived usefulness. Participants highlighted AR's potential to visualize complex concepts and engage students in experiential learning. However, limitations in 3D content availability and customization highlighted the need for integrated 3D content creation and editing tools specifically tailored to agroecology. To answer RQ7, the study conducted a systematic literature review of 60 studies to identify current accessibility evaluation methods in AR. Most evaluations employed task-based scenarios, utilizing metrics such as time on task, error rate, and user satisfaction. The study's own evaluation validated that DHH users could independently author AR content using the developed tool. In contrast, BLV users could navigate the authoring functionalities but were unable to fully author AR content, indicating that further design improvements and assistive functionalities are required for full inclusion. Methodologically, this study contributes a novel integration of DBR and Agile Scrum for inclusive educational technology design. This hybrid framework facilitated rapid prototyping, iterative refinement, and participatory co-design, and is recommended for broader application in accessibility-focused educational innovation. Future research should document and validate this methodological approach across additional contexts and user groups. The study makes the following contributions: (1) provision of an open-source, extensible AR authoring interfaces and codebase for public use; (2) improved AR accessibility for AR for DHH and BLV users; (3) development of modular architectural and algorithmic solutions to enable multimodal accessibility; (4) empirical validation of AR’s pedagogical value in agroecology education; and (5) identification of optimal open-source licensing models for collaborative educational software development. In sum, the findings demonstrate that an accessible, open-source AR authoring tool can empower diverse educators, including those with disabilities, to create inclusive and contextually relevant learning experiences. The research affirms the importance of universal design, participatory development, and modularity in educational technology design and concludes with strategic recommendations: integrating AI-assisted 3D content generation, expanding accessibility to additional user groups, and establishing communities of practice to support sustainable AR content development in agroecology.