• Version
• Download 11
• File Size 305.97 KB
• Download

Intelligent Edge Testing: Ensuring Performance and Reliability in AR/VR Devices with Edge AI

Authors:

Name: Santosh Kumar Jawalkar,

Email: santoshjawalkar92@gmail.com,

State/ Country: Texas, USA.

Abstract:

AR and VR devices become more effective with Edge AI integration resulting in transformative experiences. This technology provides quick data processing combined with enhanced user interaction along with independent operations without needing cloud platforms. AR/VR platforms deliver unsustainable user experience since cloud-based systems produce slow processing times and network dependence delays the user experience. Fast response times are attainable through the direct processing of AI workloads by implementing Edge AI technology onto edge devices. Better first-person shooter performance together with improved virtual environment responsiveness are additional benefits of this technology. Edge AI deployment in AR/VR technology also generates new challenges because of power usage problems alongside thermal issues and add Edge-to-cloud communication requirements. The proposed research introduces an edge testing framework that assesses the performance along with reliability and scalability aspects in Edge AI-powered AR/VR systems.

Auto latency measurement and online evaluation of AI processing speed and networking stability assessment make up the proposed testing infrastructure. The evaluation measured motion-to-photon latency values in combination with jitter performance alongside FPS stability along with AI model inference speed under multiple

conditions for Edge AI assessment. The investigation evaluated how edge-cloud synchronization performs while emphasizing the influences of network congestion with related bandwidth restrictions along with update delay durations in real-time AR/VR delivery. The research uses industry standard tools including Unity Profiler, OpenXR, TensorRT and Wireshark to finish a complete performance evaluation of Edge AI-enabled AR/VR applications.

Experimental evaluations show Edge AI delivers motion-to-photon latency below the acceptable level where results stay at 14ms on average. The processing time on edge devices using AI inference reached minimum levels of 8ms thus enabling real-time gesture detection and object identification. Tests exposed two main difficulties consisting of heat-related restrictions and elevated power usage when maintaining AI data processing operations. The network reliability testing confirmed packet loss together with jitter fluctuations persist in cloud-dependent applications until appropriate benchmarks for adaptive bandwidth management and real-time synchronization could be achieved.

The study demonstrates how Edge AI works to improve AR/VR applications through improved functionality alongside better performance speed and higher scalability function without requiring cloud resources. Moving forward the technology requires better dynamical resource distribution together with AI-based anomaly detectors as well as device optimization which tackles both heat generation and energy usage issues. Standard benchmarking metrics for Edge AI AR/VR applications must be developed to guarantee consistent testing results across all components of hardware as well as network environments. Edge AI will keep advancing future AR/VR innovations by solving existing obstacles to produce highly immersive responsive efficient virtual environments.

Keywords - Edge AI, AR/VR Performance Testing, Real-Time Latency, Cloud-Edge Communication, FPS Stability, AI Inference Optimization, Motion-to-Photon Delay, Network Jitter, Adaptive AI Models, Intelligent Edge Testing Framework.