Real-time immersive animation using IoT-enabled edge computing and AI for next-generation intelligent systems

The next-generation immersive animation with enhanced real time is in high demand because of the fast development of telepresence, extended reality, and metaverse applications. However, there are privacy concerns, low-latency connectivity, and high-fidelity immersive experiences that need to be considered when using this technology. To address the existing challenges, this paper proposes a new, unified framework called Federated Generative Motion-Rendering with Adaptive Edge-IoT Collaboration (FGMR-AEC) that brings together IoT sensing, edge intelligence, and federated generative learning, enabling low latency, privacy-preserving, high-fidelity immersive animation. In FGMR-AEC, AI-powered Generative Edge-Twin Networks (GETNs) interpret multimodal sensory input from distant IoT-based devices ranging from trajectories, facial expressions, gestures to ambient environmental surroundings. This system relies on an adaptive federated updating methodology to balance network communication efficiency and synchronization and supports model updating based on spatio-temporal data entropy. In addition, a hybrid compression-generation pipeline converts high-dimensional sensory input into edge-generated latent codes and reduces virtual network loads while keeping rendering fidelity high. Finally, a reinforcement learning scheduler will account for quality of service, minimize energy consumption, minimize latency, and effectively distribute resources among cloud servers, edge nodes, and IoT devices in order to optimize overall system performance. This paper introduces FGMR-AEC, a new framework for real-time, immersive animation that leverages IoT-enabled edge computing and AI-driven generative models. FGMR-AEC combines adaptive edge-device cooperation with privacy-preserving motion synthesis to enable low-latency, high-fidelity rendering across dispersed IoT nodes, unlike FL or edge animation methods. Experimental results show a 42% reduction in latency and a 0.87 improvement in structural similarity index (SSIM) over state-of-the-art edge animation frameworks, demonstrating its efficacy for next-generation intelligent systems. AEC reduces end-to-end latency by 48% and bandwidth efficiency by 60% compared to cloud-only baselines, according to experiments. The model maintains animation quality at a 95% structural similarity index, while privacy analysis ensures that pure motion and biometric data remain in the IoT layer. For the first time, FGMR-AEC combines IoT, edge intelligence, and AI-driven generative modeling to create real-time federated immersive animation.