Urban intelligent assistant on the example of the escalator passenger safety management at the subway stations.

Intelligent assistants often struggle with the complexity of spatiotemporal models used for understanding objects and environments. The construction and usage of such models demand significant computational resources. This article introduces a novel multilevel spatiotemporal model and a computationally efficient construction method. To facilitate model construction on different levels, we employ a meta-mining technique. Furthermore, the proposed model is specifically designed to excel in foggy environments. As a practical application, we develop an intelligent assistant focused on enhancing subway passenger safety. We present case examples involving jammed objects, such as shoes, in escalator combs. Our results demonstrate the effectiveness of the proposed model and method. Specifically, the accuracy of breakdown detection has improved by 10% compared to existing information systems used in subways. Moreover, the time required to build a spatiotemporal model is reduced by 2.3 times, further highlighting the efficiency of our approach. Our research offers a promising solution for intelligent assistants dealing with complex spatiotemporal modeling, with practical applications in ensuring subway passenger safety.