The railway inspection and monitoring technology mainly covers infrastructure inspection and monitoring, mobile equipment inspection and monitoring, and comprehensive inspection. In the field of infrastructure inspection and monitoring, in recent years, advanced means such as laser, ultrasonic, wireless sensor network, and drone have been widely used for the inspection of tracks, tunnels, bridges, OCSs, etc. In the field of mobile equipment inspection and monitoring, the sensor technology is mainly used for real-time monitoring of key components of rolling stock such as wheels, axles and braking devices. In the field of comprehensive inspection, faster and more automated comprehensive inspection trains are developed for the comprehensive inspection of infrastructure and mobile equipment such as tracks, OCSs, and rolling stock. In general, the railway inspection and monitoring technology shows the following development trends.
(1) High speed, low cost and high precision
As laser, ultrasonic, wireless sensor network and other technologies are developed and applied, railway infrastructure inspection focuses more on efficiency based on reliability, in a bid to improve inspection precision and speed while reducing cost.
Japan has developed a new type of rail flaw detection device for track inspection based on ultrasonic detection technology , which can achieve a flaw detection speed of 1000mm/s; its bridge detection device "U Doppler Radar" featuring wide measuring range and high accuracy is used to detect bridge deflection, natural frequency of microvibration and other items as train passes; the crack width measuring instrument for tunnel inspection can accurately measure a ±5mm deformation or a 5000-micron crack. Germany has developed the contact wire ranging system (CRS) and the contact wire wear monitoring system (WWS), which show excellent resolution when train runs at high speed. The Network Rail of UK has the New Measurement Train (NMT) with a maximum operation speed of 200km/h; it can achieve accuracy within 1 meter for all data location markers and identify wears shorter than 0.3mm on contact wires.
(2) Automation and unmanned operation
Traditional inspection and monitoring technology relies on a large number of manual operations, resulting in high cost, low efficiency and various limitations and safety hazards. With the development of modern technology, wireless sensor network, drone and other advanced technologies are applied for railway inspection and monitoring. Automation and unmanned operation have become one of the major directions in foreign countries for railway inspection and monitoring.
Railway companies in the United States use drones to collect information including images, videos, and LIDAR survey data as auxiliary inspection for rails and bridges. Drones can work efficiently without interrupting traffic operation and acquire pictures from locations inaccessible for people. Japan has developed a tunnel damage automatic monitoring system based on wireless sensor network, which uses inspection instruments to collect and transmit health data of tunnel lining and carry out long-term remote automatic monitoring of tunnels. Germany has developed a centralized remote monitoring device, which can continuously collect various data from all types of locomotives or trains and share them with customers and maintenance partners in a real time manner. South Africa has developed an automatic inspection vehicle (SID) equipped with laser scanners, multispectral cameras and other equipment, which can automatically detect obstacles, track conditions and OCS, etc., and transmit the inspection results to the command center.
(3)Operation interruption-free
To avoid operation interference and enhance carrying capacity of railway lines, two solutions are adopted in foreign countries to conduct inspection and monitoring. One is to install inspection system on commercial trains to monitor train status and then supervise conditions of vehicles and tracks. The other is to install sensors on tracks and receivers on trains, to obtain data as the train passes.
Australia uses instrumented revenue vehicles (IRVs) to continuously monitor track conditions and vehicle dynamic property, which helps to reduce faults and accidents. Japan has developed a device using commercial trains to collect monitoring data by trackside installing a wireless sensing device composed of transmitter and sensor and a receiver onboard the train, which can receive multi-group of data in a short time. Track IQ, a subsidiary of Westinghouse Air Brake, has developed the wheel condition monitor (WCM) with modular sensors clamping onto rails without changing track structure. With easy maintenance and quick installation, it won’t interfere with normal traffic operation or routine track maintenance.
(4) Integration of inspection & monitoring and fault diagnosis
Thank to wide application of sensor technology in detection and monitoring and rapid development of big data and AI, it is possible to achieve intelligent diagnosis and prediction of infrastructure status. In recent years, foreign railways have been carrying out research on the system integrating inspection and monitoring and fault diagnosis to promote the evolvement of traditional "periodic repair" to the "repair according to condition".
Shift2Rail, a railway technology, research and development collaboration program in Europe, launched the MOMIT project, using technologies such as earth observation satellites and RPAS-borne sensors to collect information of railway infrastructure status and support repair according to condition and intelligent assets management via data analysis. The EMT of the UK has developed track fault prediction system based on machine learning and big data processing technologies, which acquires and transmits all information about railway infrastructure via IoT sensors, and utilizes artificial intelligence (AI) techniques to process and assess the data and predict possible faults on a 3D virtual model of stations and tracks. Japan has developed tunnel damage software based on deep learning technology, which can quickly process tunnel lining images and accurately discover and mark tunnel cracks with a width exceeding 0.5mm. Railways in Switzerland have developed a train monitoring network to monitor the condition of rolling stock through data measured by train monitoring devices, helping predict the future development of rolling stock and schedule the best maintenance time.
