ANALYSIS OF ARTIFICIAL INTELLIGENCE AND IOT APPROACHES TO INTELLIGENT MONITORING AND AUTOMATIC CONTROL OF SMART CITY TECHNOLOGICAL PROCESSES

The article analyzes artificial intelligence and IoT approaches to intelligent monitoring and automatic control of smart city technological processes within a single closed control loop. The loop is divided into four layers: sensing and communication, edge, cloud and control. Perception methods, edge computing and control algorithms are compared by five criteria: perception accuracy, control quality, computational demand, data demand and privacy. The analytical results of the article are the closed-loop model, the comparison of approaches and three main open problems: propagation of perception error into control, resource limits of edge devices and privacy requirements. For comparison, an earlier published study of the author is cited: the measure-based method reaches 31.7 AP on AI-TOD-v2. These results define the directions of further research.