Indoor Scene Construction Technology Based on 6G Virtual Simulation and CPS

Our study investigates using a unified framework for indoor scene creation in industrial settings that combines Cyber-Physical Systems (CPS) with 6G technologies. This research aims to improve automation and real-time interaction in intricate industrial environments by utilising the superior capabilities of 6G and CPS. We provide a case study of a manufacturing facility to demonstrate how our method makes space optimisation easier, boosts operational effectiveness, and strengthens safety protocols. This case study is an excellent example of the potential and real-world advantages of using cutting-edge technologies in industrial settings. This study delves into the challenge of failure prediction in process sectors that use intelligent and autonomous cyber-physical systems (CPS) in a 6G setting. This aligns with the latest developments in Industry 4.0 and the IIoT. Specifically, we developed a full-stack deep learning approach that used massive amounts of real-time sensory data collected from wireless sensors in a chemical plant. To start, while working with unbalanced time-series data, a unique recursive architecture is proposed that uses several lookback inputs to make an initial forecast using autoregression. During this method, a new learning algorithm called "Recursive Gradient Descent (RGD)" is developed for the proposed architecture to reduce the cumulative prediction uncertainties. Afterwards, a multi-class classification model using temporal convolutions across many channels with a decay effect is proposed to detect and localise the root causes of failure. Because of its exceptional ability to reduce prediction uncertainties accumulated across numerous prediction stages, the entire network is termed the Cumulative Uncertainty Reduction Network with Bayesian Neural Network (CURN-BNN). Results show that CURN-BNN outperforms the state-of-the-art approaches, especially regarding recall for fault prediction and fault type categorisation accuracy.