DIGITAL LIBRARY
AUGMENTED REALITY AS VISUAL SCAFFOLDING FOR 2D-TO-3D CONSTRUCTION TASKS: A ROBOTICS-INTEGRATED STEM EDUCATION CASE STUDY
Braude College of Engineering (ISRAEL)
About this paper:
Appears in: EDULEARN26 Proceedings
Publication year: 2026
Article: 0855 (abstract only)
ISBN: 978-84-09-88444-5
ISSN: 2340-1117
doi: 10.21125/edulearn.2026.0855
Conference name: 18th International Conference on Education and New Learning Technologies
Dates: 29 June-1 July, 2026
Location: Palma, Spain
Abstract:
Hands-on STEM learning often involves constructing three-dimensional (3D) models from two-dimensional (2D) instructions, a transition that can be challenging for novice learners with limited spatial visualization skills. In educational robotics contexts, assembly guidance is typically provided through static 2D diagrams, while the constructed artifacts are inherently 3D. Translating these representations into accurate 3D constructions can be cognitively demanding and may hinder spatial learning during hands-on STEM activities. These difficulties can lead to assembly errors, learner frustration, and increased reliance on instructor support, potentially affecting students’ engagement, motivation, and perceived self-efficacy during construction tasks. Although robotics assembly provides a clear example of this challenge, similar difficulties arise in various technology-enhanced learning environments where learners must translate 2D procedural representations into 3D physical actions.

Augmented Reality (AR) offers new possibilities for supporting learners in tasks that require interpreting spatial relationships between 2D representations and 3D constructions. This paper presents AR-semble, a mobile AR prototype developed as a pilot proof-of-concept for supporting students during robotics assembly tasks. Rather than replacing traditional instructions, the system augments 2D assembly guidance by overlaying aligned 3D visual support that demonstrates the orientation and placement of components during each construction step. The prototype illustrates how multi-step assembly processes can be supported through AR-based visual scaffolding within existing instructional workflows. The contribution of this work lies in presenting a design-oriented perspective in which AR functions as instructional visual scaffolding for learners who struggle to translate 2D representations into 3D constructions.

The design of AR-semble was informed by consultations with three educational robotics experts with extensive experience in youth STEM programs. These consultations highlighted recurring classroom challenges, including difficulties in identifying similar components, interpreting spatial orientation from diagrams, and progressing at different paces during group assembly activities. In response, the prototype incorporates step-by-step visual guidance, simple learner-controlled navigation, gamification mechanisms, and supportive design features intended to sustain engagement during construction tasks.

From a technical perspective, AR-semble was developed using Unity with the Vuforia AR tracking library, using marker-based AR to align digital 3D models with 2D assembly instructions. The 3D assets are authored manually through a LeoCAD-to-Unity pipeline, a constraint acknowledged as a scalability consideration and a direction for future semi-automated authoring work. This implementation demonstrates a flexible development approach that can support the creation of similar AR-based learning applications for a variety of hands-on STEM construction activities.

Importantly, the system is intended to support learners who struggle with spatial interpretation rather than replace traditional construction practices that help develop spatial reasoning skills. By offering optional visual scaffolding, AR-semble aims to support more inclusive participation and personalized learning pathways during hands-on STEM construction tasks.
Keywords:
Augmented Reality, Robotics-Integrated STEM Education, Visual Scaffolding.