**2. Literature review on track condition monitoring based on in-serve train vibration**

#### **2.1 Track condition monitoring from an in-service vehicle**

Track maintenance and management on the railway is based on measured data of track displacement. However, track displacement measurement requires expensive equipment such as track inspection vehicles and measuring devices. A more economical management method is required, especially for regional railways.

Other methods of track management exist in addition to displacement measurement, e.g., train vibration inspection using in-service vehicle vibration measurement [6, 7]. Many studies on track monitoring systems using in-service vehicle vibration measurement with on-board sensing devices were conducted both in Japan and abroad.

#### **2.2 Axle-box-mounted sensors**

Chen *et al.*, Karis *et al.*, and Tsai *et al.* analyzed the relation between axle-box accelerations and railway defects or irregularities [8–10].

Sun *et al.* proposed an on-board detection technique for longitudinal track irregularity that can be applied to commercial high-speed trains. The acceleration of the axle-box of the high-speed train was evaluated [11].

Chudzikiewicz *et al.* demonstrated the possibilities of estimating the track condition using axle-boxes and car-bodies motions described by acceleration signals. They presented the preliminary investigation on the test track and supervised runs on Polish Railway Lines of an Electric Multiple Unit (EMU-ED74), [12].

#### **2.3 Bogie-mounted sensors**

Some types of track faults were detected by measuring the acceleration of bogies. Weston *et al.* demonstrated track irregularity monitoring by using bogie-mounted sensors [13, 14].

Malekjafarian *et al.* investigated the use of drive-by train measurements for railway track monitoring on the Dublin-Belfast line with an in-service Irish Rail train. The measurements were taken with accelerometers and a global positioning system. They used the train bogie accelerations [15].

#### **2.4 Car-body-mounted sensors**

Tsunashima *et al.* developed a system to identify track faults by using accelerometers and GNSS placed on the car-body of in-service vehicles [1–5].

*Track Condition Monitoring Based on In-Service Train Vibration Data Using Smartphones DOI: http://dx.doi.org/10.5772/intechopen.111703*

Bai *et al.* used low-cost accelerometers that were placed on or attached to the floors of operating trains for analyzing track quality [16].

A track condition monitoring based on the bogie and car-body acceleration measurements was presented and verified in Shang Hai metro Line 1 [17].

Balouchi presented a cab-based track monitoring system developed in the UK. They presented through comparison of vibration response from sites with known defects and outputs from Network Rail's New Measurement Train (NMT). Good agreement was reported for track faults in relation to vertical and lateral alignment and dip faults [18].

#### **2.5 Signal processing**

To extract a signal on faulty tracks from measured vehicle vibration, several techniques using nonmodel-based and model-based method were proposed.

Tsunashima *et al.* proposed a nonmodel-based technique using time-frequency analysis [4]. In this paper, detection performance using continuous wavelet transform (CWT) and Hilbert-Huang transform (HHT) were compared for identifying track faults from car-body vibration. They showed that track fault features can be identified in the time-frequency plane based on the analysis of simulation studies and field tests.

A Kalman filter-based method to estimate the track geometry of Shinkansen tracks from car-body motions was proposed [19]. The proposed Kalman filter-based estimation technique was modified and applied for conventional railways [20].

Tsunashima proposed a classifier based on machine learning techniques for identifying track faults automatically from measured car-body vibration [3]. It is shown that the degradation of track can be classified in the feature space consists of car-body vibration RMS.

A new method for automatically classifying the type and degradation level of track faults using a convolutional neural network (CNN) by imaging car-body acceleration on a time-frequency plane by continuous wavelet transform [5].

#### **2.6 Smartphones-based system**

Chellaswamy *et al.* proposed a method for monitoring the irregularities in railway tracks by updating the status of the tracks in the cloud. The IoT based Railway Track Monitoring System (IoT-RMS) is proposed for monitoring the health of the railway track [21].

Rodríguez *et al.* presents the use of mobile applications to assess the quality and comfort of a railway section track (narrow gauge) in northern Spain [22].

Cong *et al.* proposed an approach for using the smartphone as a sensing platform to obtain real-time data on vehicle acceleration, velocity, and location for monitoring the track condition during subway rail transit in China [23].

Paixão *et al.* presented an approach to use smartphones to perform constant acceleration measurements inside in-service trains to complement the assessment of the structural performance and geometrical degradation of the tracks. To demonstrate the applicability of smartphone's sensing capabilities for on-board railway track monitoring, they evaluated the accelerations inside the car-body of the Portuguese Alfa Pendular passenger train [24].
