Autonomous Drive-by Monitoring Technologies for Developing Smart Transportation Infrastructure Systems

我们现代的宏伟事业之一是建设智能、可持续、有弹性的城市，在为居民提供高水平舒适度的同时，能够高效地管理和运营。实现这一目标的道路本质上是复杂和多维的，它需要从根本上改变我们基础设施的管理。开发一个容量、效率和弹性更高的智能交通基础设施系统，并利用信息技术进行有效的运营和维护，是这项事业的关键一步。 智能基础设施的一个重要组成部分是持续感知和监测。为了促进智能和可持续交通基础设施的发展，本研究提出了一种新的范式，利用众包数据对全球交通基础设施系统的人口进行持续和自动化的监测。由于传感技术和物联网的快速发展，我们交通基础设施上使用的车辆以及车内乘客的智能手机通常都配备了各种传感器，如加速计、摄像头和GPS。通过拟议的研究，将开发利用众包车辆内现有传感器的方法，对交通基础设施人口进行自动监测。 在该框架中，每辆车作为移动感知节点进行数据采集。一旦收集到数据，将使用车载处理器对其进行预处理，并将提取的特征传输到远程数据库，以供进一步分析和决策。车载智能手机中嵌入的加速度计将被用来捕捉桥梁和车辆的耦合振动。还将开发先进的信号处理技术，使用来自大量车辆上的智能手机的振动数据，以自动方式识别桥梁人口上的损伤的存在、位置和严重程度。此外，拟议的研究将开发涉及深度学习的有效图像处理技术，以便使用安装在车辆上的摄像头(例如后备摄像头和仪表板摄像头)实时高精度地自动检测道路或桥梁表面的缺陷。 据研究人员所知，这个框架是世界上第一个这样的框架，将带来一种颠覆性的技术和一个新的研究领域。拟议的系统有可能连续、实时地自动监测交通基础设施系统的人口。拟议研究的预期结果将帮助加拿大的基础设施所有者提高现有基础设施的安全性和可持续性，并为未来的智能城市创造可持续的智能基础设施。在未来，建议的框架可以扩展到使用现有传感器的智能互联自动驾驶车辆。

One of the grand undertakings of our modern age is to build smart, sustainable, and resilient cities that can be managed and operated efficiently while providing residents with a high level of comfort. The path to achieving this is inherently complex and multi-dimensional, and it requires a fundamental shift in the management of our infrastructure. Developing a smart transportation infrastructure system with higher capacity, efficiency, and resilience, which can be operated and maintained effectively using information technologies, is a key step in this undertaking. One important component of smart infrastructure is continuous sensing and monitoring. To contribute to the development of smart and sustainable transportation infrastructure, this research proposes a novel paradigm for continuous and automated monitoring of populations of transportation infrastructure systems globally using data from crowdsourcing. Owing to rapid advances in sensing technologies and IoT, the vehicles used on our transportation infrastructure, as well as the smartphones of the passengers in vehicles, are usually equipped with various sensors such as accelerometers, cameras, and GPS. Through the proposed research, methods will be developed for automated monitoring of populations of transportation infrastructure capitalizing on the existing sensors within crowdsourced vehicles. In this framework, each vehicle serves as a mobile sensing node for data collection. Once the data has been collected, it will be pre-processed using on-board processors, and extracted features will be transmitted to a remote database for further analysis and decision making. Accelerometers embedded in the smartphones in the vehicles will be used to capture the coupled vibration of bridges and vehicles. Advanced signal processing techniques will also be developed for identifying the existence, location, and severity of the damage on populations of bridges in an automated fashion using the vibration data from smartphones in a large number of vehicles. Additionally, the proposed research will develop effective image processing techniques involving deep learning for automated detection of defects on road or bridge deck surfaces with high accuracy in real time using the cameras installed in vehicles (e.g., backup and dash cameras). To the best of the researcher's knowledge, this framework is the first of its kind in the world, and will lead to a disruptive technology and a new area of research. The proposed system has the potential to continuously monitor populations of transportation infrastructure systems automatically in real time. Expected outcomes of the proposed research will help Canada's infrastructure owners to increase the safety and sustainability of existing infrastructure, and to create sustainable smart infrastructure for the smart cities of the future. In the future, the proposed framework can be extended to use with smart and connected autonomous vehicles employing existing sensors.