Towards proactive maintenance of buried infrastructure with cloud-based sensing and predictive analytics

包括水、废水和雨水管网在内的地下基础设施的状况对人们的生活质量非常重要，并具有强烈的环境、经济和社会影响。最近加拿大各地的一项市政地下基础设施调查发现，水污染、安全和供水系统的状况是这个国家面临的最大问题。城市的漏水正在浪费数百万美元的税款。公用事业公司正在主要通过被动的方法来解决这些问题。需要一种基于性能的方法来确定各个供水管道是否达到了所需的服务水平。然而，由于技术和财务方面的限制，这些隐藏资产的业绩管理历来没有进行过。主动、统一和系统地维护埋在地下的基础设施将是更有效率和更可取的。 为了实现对地下基础设施的主动维护，必须从管道中收集所有相关数据和信息，以确定未来的管道状况。目前，进行现场监测和无损检测来收集数据。然而，这些数据以不同的格式保存，并存储在不同的位置。由于这些数据彼此之间没有正确匹配和对齐，因此只能使用孤立的数据来构建预测模型。基于积累的检测数据的综合预测可用于开发下一代主动维护。这项研究计划将物联网、云计算和高级机器学习等现代信息和通信技术融入地下基础设施管理，以填补技术空白。 该研究计划的长期目标是推进状况评估技术和预测分析，以通过主动维护实现地下基础设施的可持续性。短期内，该研究计划将开发一个基于云的传感框架，该框架将实现集中化的数据收集、存储和管理，并开发预测性分析，从大数据中获取可操作的信息。 主动维护使自来水公司的预防和纠正措施能够得到适当和有效的安排。因此，给加拿大地下基础设施带来的影响包括更高的系统可靠性，降低管道故障的成本，最大限度地减少维护时间，以及优化维护间隔。拟议的研究还将为HQP提供一个独特的跨学科培训机会。HQP将获得不止一个学科的深入知识和理解，例如电气、土木和机械工程，并能够解决实际问题，而不受专业界限的限制。

The condition of buried infrastructure including water, wastewater, and storm water networks is of great importance to people's quality of life and has strong environmental, economic and social implications. A recent municipal buried infrastructure survey across Canada identified water pollution, safety and state of water supply systems as the greatest issues to this country. City water leaks are wasting millions of tax dollars. Utilities are addressing these issues primarily with a reactive approach. A performance-based approach is needed to identify if individual water pipelines meet the required level of service. However, performance-based management of these buried assets has historically not been performed due to technical and financial limitations. A proactive, uniform, and systematic approach for maintenance of buried infrastructure would be more efficient and preferred. To achieve the proactive maintenance of buried infrastructure, it is essential to collect all the relevant data and information from the pipelines to determine future pipe condition. Currently, in-situ monitoring as well as non-destructive inspection are performed to collect the data. However, these data are saved in different formats and stored in varied places. As these data are not properly matched and aligned to each other, predictive models can only be built with isolated data. A comprehensive prediction based on accumulated inspection data can be used to develop next-generation proactive maintenance. This research program will fill the technical gaps by incorporating modern information and communications technologies, e.g. Internet of Things, cloud computing, and advanced machine learning, into the buried infrastructure management. The long-term objective of the research program is to advance condition assessment technologies and predictive analytics to achieve the sustainability of buried infrastructure through proactive maintenance. In the short-term, the research program will develop a cloud-based sensing framework, which will enable centralized data collection, storage, and management, and develop predictive analytics to derive actionable information from big inspection data. Proactive maintenance allows preventive and corrective actions from water utilities to be properly and efficiently scheduled. Thus, the impacts brought to Canada's buried infrastructure include higher system reliability, reduced costs of pipe failures, minimized time on maintenance, and optimized maintenance interval. The proposed research will also offer a unique cross-disciplinary training opportunity for HQP. The HQP will gain in-depth knowledge and understanding of more than one discipline, e.g. electrical, civil, and mechanical engineering, and be able to solve practical problems without regard to disciplinary boundaries.