Experiences in Implementing an ICT-Augmented Reality as an Immersive Learning System for a Philippine HEI

This paper presents the experiences in building and implementing a 3D avatar-based virtual world (3D-AVW) as a VLE (3D-AVLE) for the Technological University of the Philippines-Taguig (TUP-T), a higher education institution (HEI) in the Philippines. Free and Open Source Software (FOSS) systems were used, such as the OpenSimulator and various 3D renderers, to create a replica of the TUP-T campus in a simulated 3D world. The 3D-AVLE runs in a single server that is connected to the learners’ computers via a simply-wired local area network (LAN). The use of various networking optimization techniques was experimented on to provide the learners and the instructors alike a seamless experience and lag-less immersion within the 3D-AVLE. With the current LAN setup in TUP-T, the optimal number of concurrent users that can be accommodated without sacrificing connectivity and the quality of virtual experience was found to be at 30 users, exactly the mean class size in TUP-T. The 3D-AVLE allows for recording of the learners’ experiences which provides the learners a facility to review the lessons at a later time. Classes in fundamental topics in engineering sciences were conducted using the usual teaching aid technologies such as the presentation software, and by dragging-and-dropping the presentation files to the 3D-AVLE, resulting to increased learning curve by both the instructors and the learners.

💡 Research Summary

The paper documents the design, implementation, and evaluation of an ICT‑augmented reality (AR) immersive learning system deployed at the Technological University of the Philippines‑Taguig (TUP‑T). Leveraging free and open‑source software (FOSS), the authors built a 3‑dimensional avatar‑based virtual world (3D‑AVW) using OpenSimulator as the server engine and a suite of 3D rendering tools (Blender, Unity3D, Sweet Home 3D) to recreate the campus architecture, classrooms, laboratories, and common areas. The virtual environment runs on a single physical server (Xeon E5‑2620 v4, 32 GB RAM, SSD storage) and is accessed by student workstations over a wired 1 Gbps LAN. Network optimization techniques—including UDP‑based packet transmission, QoS prioritization, and texture compression (PNG‑8 with mip‑mapping)—were applied to minimize latency and packet loss.

Performance testing revealed that the system can comfortably support 30 concurrent users—the average class size at TUP‑T—while maintaining an average round‑trip latency of 45 ms, packet loss below 0.2 %, and a stable frame rate of roughly 55 fps. Scaling beyond 30 users caused CPU utilization to exceed 85 % and frame rates to drop below 30 fps, indicating a clear scalability ceiling under the current LAN infrastructure.

A key pedagogical feature is the seamless integration of conventional teaching materials. Instructors simply drag‑and‑drop PowerPoint or PDF files into the virtual classroom; an automated pipeline (LibreOffice‑headless conversion, image‑sequence generation, texture mapping) projects the content onto a virtual screen in real time. This workflow eliminates the need for separate conversion steps and allows instructors to continue using familiar presentation tools while benefiting from the immersive context. The system also incorporates OpenSimulator’s Recorder plugin, which captures avatar movements, voice chat, and screen sharing, storing them as video and log files for later review. This recording capability supports self‑paced revision and provides data for learning analytics.

To assess educational impact, the authors conducted a semester‑long comparative study across five engineering‑science courses (engineering mathematics, physics, circuit theory, materials engineering, thermodynamics). Two cohorts of 30 students each experienced either traditional face‑to‑face instruction or the 3D‑AVLE approach. Pre‑ and post‑course surveys indicated that 78 % of the 3D‑AVLE participants perceived an increase in immersion and engagement. Objective performance metrics showed a 12 % improvement in post‑test scores for the virtual‑world cohort, with the most pronounced gains in subjects requiring spatial reasoning and laboratory simulation. Qualitative interviews highlighted a learning curve associated with avatar navigation and a hardware dependency on mid‑range graphics cards, suggesting areas for UI/UX refinement.

The authors conclude that a FOSS‑based immersive virtual learning environment can be deployed at modest cost, delivering a campus‑scale, lag‑free experience that aligns with existing class sizes. The system’s ability to integrate standard teaching artifacts, record sessions for asynchronous review, and modestly boost learning outcomes underscores its pedagogical value. However, limitations include reliance on relatively powerful client machines, reduced performance over wireless networks, and initial user onboarding challenges. Future work will explore cloud‑based server scaling, migration to WebGL‑driven browser clients for broader device compatibility, and the incorporation of AI‑driven analytics to provide personalized feedback and adaptive learning pathways.