Monitoring Systems Design with Real Time Interactive 3D and Artificial Intelligence

The paper proposes the topic of Environmental Artificial Intelligence i.e., Artificial Intelligence approaches, based on the use of natural language, applied to architecture to support the design of systems to control the progress of degenerative states and to test their functionality through simulations with real-time interactive 3D models. Beginning with multiscalar heritage digitization techniques, the study explores the possibilities of integrating 3D models, annotated semantically, with AI for testing surface degradation progress monitoring systems. The monitoring system is based on the installation of RGB and, in the future, thermal cameras that acquire data streamed to them. Before installing such sensors in situ, their operation can be verified by taking advantage of a module in Unreal Engine 5 that simulates this streaming. The stream is hooked up to an AI group that has the task of analyzing the streaming images, recognizing if there are anomalies in relevant parts by cross-referencing the information with the semantic mapping of the digital model, and sending a signal to a dialogue system that has the task of alerting, in natural language, the user. The user can explore the 3D space in real time to check areas where there has been a significant change.