Multi-sensor condition monitoring for predictive maintenance of rail infrastructure using optical fibre sensors (OptRail)

铁路基础设施的维护是铁路行业的一项主要成本，每年花费超过**10亿GB**。占Network Rail支出的**18%**。它也是计划内和计划外维护操作造成网络中断的主要原因。铁路使用量和对铁路服务的需求正在迅速增长，这使得铁路上的交通流量越来越大，对维护的要求也越来越高。**预测性维护**使用轨道及其状况的数据和模型来估计剩余使用寿命，并在需要的地方进行目标维护，以减少不必要的维护，防止在发生故障时进行计划外的反应性维护。节省的资金可能是巨大的。Network Rail已经确定，达到世界级的预测性、基于风险的维护策略可以带来以下好处：*25%-35%*降低*维护成本*70%+*减少*服务故障次数**35%-45%**故障后的停机时间*20%+提高**员工生产力*更少的计划外、反应性干预，从而增强员工安全。实施这些策略需要准确、详细和最新的铁路基础设施状态和条件知识。这是使用传统的人工检查和用于监测轨道和基础设施状况的新技术无法获得的，而自动化数据采集、分析和维护规划将是实现这些好处的关键。该项目将开发一种新型的轨道维护计划自动化系统，使用为在石油和天然气行业遇到的恶劣环境中使用而开发的**光纤传感器**的独特组合，以及新一代**物联网(IoT)**通信技术和**人工智能**，以提供**自动决策支持**工具，用于基于对轨道状况的连续实时监测来优化维护计划。该系统将补充和连接现有的**数字检测技术**，如测量列车，满足对连续、**实时监测**和基于模型的预测和优化的需求。传感器系统设计的核心是可用性和低成本。OptRail将被开发为一个模块化系统，将传感器元件和接口构建到一个坚固的封装中，该封装可以可靠地粘合到轨道上或嵌入结构中，只需最少的工作量和停机时间。

Maintenance of rail infrastructure is a major cost to the rail industry, costing over **£1billion p.a**. in the UK and representing **18% of Network Rail's expenditure**. It is also a major source of network disruption from both planned and unplanned maintenance operations. Rail usage and demand for rail services is increasing rapidly, placing more traffic on the rails and increasing requirements for maintenance.**Predictive maintenance** uses data and models of the rail track and its condition to estimate remaining useful life and target maintenance where it is needed to reduce unnecessary maintenance prevent unplanned reactive maintenance in case of failure. The savings can be huge. Network Rail have identified that reaching world-class predictive, risk-based maintenance strategies could deliver the following benefits:* **25-35%** **reduction** in maintenance costs* **70%+** **reduction** in the number of service failures* **35-45% reduction** in down time following failure* **20%+ increase** in workforce productivity* fewer unplanned, reactive interventions delivering enhanced workforce safety.Implementation of these strategies requires accurate, detailed and up to date knowledge of the state and condition of the railway infrastructure. This cannot be obtained using traditional manual inspection and new technologies for monitoring track and infrastructure condition, and automating data acquisition, analysis and maintenance planning will be essential to deliver these benefits.This project will develop a novel automated system for planning of track maintenance, using a unique combination of **optical fibre sensors** developed for use in harsh environments encountered in the oil and gas industry, coupled with new generation **Internet of Things (IoT)** communications technology and **artificial intelligence** to provide **automated decision support** tools for optimisation of maintenance programmes based on continuous real-time monitoring of track condition. The system will complement and interface with existing **digital inspection technologies** such as measurement trains, filling a need for continuous, **real-time monitoring** and model-based prognostics and optimisation. Central to the design of the sensor system are usability and low-cost. OptRail will be developed as a modular system with the sensor elements and interfaces built into a robust package that can be reliably bonded to tracks or embedded in structures with minimum effort and downtime.