Silva Andre A F, Porto Adao J S, Belo Bruno M C, Cesar Cecilia A C
Computer Science Division, Aeronautics Institute of Technology, São José dos Campos 12228-900, Brazil.
Sensors (Basel). 2024 Dec 13;24(24):7956. doi: 10.3390/s24247956.
Current technologies could potentially solve many of the urban problems in today's cities. Many cities already possess cameras, drones, thermometers, pollution air gauges, and other sensors. However, most of these have been designated for use in individual domains within City Hall, creating a maze of individual data domains that cannot connect to each other. This jumble of domains and stakeholders prevents collaboration and transparency. Cities need an integrated system in which data and dashboards can be shared by city administrators to better deal with urban problems that involve several sectors and to improve oversight. This paper presents a model of an integrative system to manage classes of problems within one administrative municipal domain. Our model contains the cyber-physical system's elements: the physical object, the sensors and electronic devices attached to it, a database of collected problems, code running on the devices or remotely, and the human. We tested the model by using it on the recurring problem of potholes in city streets. An AI model for identifying potholes was integrated into applications available to citizens and operators so that they can feed the municipal system with images and the locations of potholes using their cell phone camera. Preliminary results indicate that these sensors can detect potholes with an accuracy of 91% and 99%, depending on the detection equipment used. In addition, the dashboards provide the manager and the citizen with a transparent view of the problems' progress and support for their correct address.
当前的技术有可能解决当今城市中的许多城市问题。许多城市已经配备了摄像头、无人机、温度计、污染空气测量仪和其他传感器。然而，其中大多数已被指定用于市政厅内的各个领域，形成了一个无法相互连接的单个数据领域迷宫。这种领域和利益相关者的混乱局面阻碍了协作和透明度。城市需要一个集成系统，城市管理人员可以在其中共享数据和仪表板，以更好地处理涉及多个部门的城市问题并加强监督。本文提出了一个集成系统模型，用于管理一个行政市政领域内的各类问题。我们的模型包含网络物理系统的元素：物理对象、连接到它的传感器和电子设备、收集到的问题的数据库、在设备上或远程运行的代码以及人。我们通过将该模型应用于城市街道上反复出现的坑洼问题对其进行了测试。一个用于识别坑洼的人工智能模型被集成到市民和操作员可用的应用程序中，这样他们就可以使用手机摄像头向市政系统提供坑洼的图像和位置。初步结果表明，根据所使用的检测设备不同，这些传感器检测坑洼的准确率分别为91%和99%。此外，仪表板为管理人员和市民提供了问题进展的透明视图，并支持对问题的正确处理。
