**1. Introduction**

The services, such as cargo and mass traveling using the railway infrastructure, have undergone tremendous progress in recent years due to their reliability and safety in comparison to other modes of transportation. However, there is a possibility of degradation in both the railway vehicles (trains/trams) and infrastructure (track lines and the associated hardware) due to an increase in the speed limit, especially in the high-speed railway sector. This degradation normally leads to an increase in accidents. In under and non-developed countries, these accidents are inevitable due to the high cost of maintenance requirements for railway vehicles and infrastructure. Therefore, both the rolling stock and vehicles require reliability and low-cost operational monitoring.

In all of the mentioned metrics, infrastructure and vehicle are the two subjects, the attributes of which can be measured. The conventional operational monitoring systems for the mentioned subjects are usually cost-effective, but this cost-effectiveness is at the expense of unreliability if compared to the current optical fiber sensing (OFS) system. A dedicated monitoring (either operational or traffic management) technique can be used to measure one of the mentioned subjects which is more expensive than using a single method to measure both subjects with reliability. With these factors in mind, conventional sensing techniques in the railway sector can be regarded as expensive. Other than addressing a single metric and unreliability, there are many other aspects of the conventional techniques that make these methods expensive

when compared to OFS. For example, a very high amount of electricity is consumed with the track circuits (TC) technique, or a very large number of sensors are normally installed to cover the operational monitoring of both the vehicles and infrastructure in the case of communication-based train control (CBTC).

A reliable sensing system refers to a number of factors, including the avoidance of dead sensing points, ease of installation, resistance of electromagnetic interference (EMI), and avoidance of safety hazards. Compared with the distributed optical fiber sensing (DOFS) technologies, the conventional railway-oriented sensing techniques are unreliable because there is no such method used by these techniques for measuring the distributed sensing, and hence there is a possibility of a large number of dead points leading to the provision of unreliable services. To avoid the possibility of dead points, there are chances of installing a myriad number of wireless sensors, but the energy requirements are challenging to meet the objectives. If wired sensors are installed instead, the number of power and communication lines will be very large, and the system installation may be impractical. Moreover, a long-term testing and assessment procedure of a fiber optic-based railway infrastructure is possible due to the possibility of high spatial resolution based on distributed and quasi-distributed nature of sensing. Such a distributed sensing system removes the necessity of power and data cables because it serves a dual purpose; sensing system with no active power requirement and as a communication system, thus reducing the cost tremendously for tens of kilometers distance span. Therefore, the difficulty of installing a myriad number of sensors in conventional sensing is less competitive than the OFS-based solution for both cost and reliability. EMI is another issue in electrical-based sensing systems, especially in very high voltage pantographs, which can be avoided with OFS as the light is the only signal passed through optical fibers. To avoid the difficulty of providing dedicated power and communication lines, TC provides a reasonable solution to cut off the additional power lines. However, this technique does not meet the standard safety requirements, which leads to an extreme sense of unreliability in case of avoidance of safety hazards.

Any railway system requires technology to help in the operational monitoring, train traffic management, and an additional amount of data for postaccident investigation. Operational monitoring is actually the structural health monitoring of a railway system, and it is based on the investigation of both the vehicle and its infrastructure. In the case of railway vehicle traffic management, any sensing solution may provide the instantaneous location, instantaneous speed detection, and live tracking of the vehicle, and these parameters are enough to control and manage the overall traffic. As the number of accidents is increasing in economically poor countries, the third benefit that any railway system can take from the technology is the acquisition of enough data, which can be beneficial in case of postaccident investigation, which is helpful to know the potential reason for the incident to avoid in the future. Other than manual inspection, several methods have been adopted so far to automate the operational monitoring of railway vehicles and infrastructure. These methods include CBTC, TC, wheel counters, track recording cars, and onboard operational monitoring.

The data, acquired in the case of conventional railway-based sensing systems, is discrete with a relatively large interval among spatial data samples if compared with any OFS-based technology. Any technology associated with DOFS sensing is normally distributed in nature, whereas there are also discrete sensing technologies in OFS such as fiber Bragg gratings (FBGs) and interferometry-based sensing solutions. However, there is a possibility of alleviating the distance between adjacent point sensors in the

case of discrete or quasi-distributed OFS in a handful of ways if compared to the conventional sensing solution. Therefore, acquiring data with minimal distance among spatial locations is possible with OFS, and this large number of data has numerous other benefits.

A discrete sensing with a relatively large interval among sensing points may help in the case of operational monitoring but in the case of traffic management and postaccident analysis, this is surely a poor solution. The reason that continuous sensing or discrete sensing with relatively short intervals among sensing points well suits traffic management and postaccidental analysis is that data is critical at every possible point in these applications. For example, instantaneous speed and position determination are critical in the case of traffic management due to the high speed of railway vehicles. Missing data due to dead points may create severe problems, which may lead to a train-to-train collision. Moreover, postaccidental analysis requires data separated spatially in close vicinity in order to make the investigation possibly easy. Therefore, the sensing points, located spatially in close vicinity, can be used for any of the application that falls under any of the three groups of applications, as discussed. However, keeping the reliability factor in mind one may target only the applications that fall under the category of operational monitoring if the sensing points are not closed enough. This conclusion leads to the fact that the selected OFS technologies provide an edge due to their capability of spatially close sensing points or distributed sensing nature if compared with conventional sensing technologies.

Over the last two decades, the OFS-based smart railway infrastructure has made tremendous progress in view of research and development as optical fiber is always preferred due to its lightweight, reliability, and cost-effectiveness. The additional benefits of these systems include their lightweight nature and the possibility of distributed sensing solutions. In the case of train tracks, there is no possibility of data disruption while the vehicle passes within the tunnels as it normally occurs in case of the navigation systems such as the Global Positioning System (GPS) and Global Navigation Satellite System (GLONASS), etc. Unlike the conventional sensing technologies in the railway sector, which measure only the frequency spectrum, the technologies associated with OFS not only measure the frequency spectrum but also the true phase of the time-based signals. Therefore, to prevent catastrophic failures and early failure detection, the real-time sensing nature of OFS-based systems is a better option if compared with the conventional sensing system. In short, OFS can be regarded as the only reliable and cost-effective solution for smart railways. Other than management system, operational inspection, and providing ease of postaccidental investigation, some technologies in OFS can also provide intrusion detection and trespassing monitoring.

In summary, OFS-based sensing systems are not only capable of targeting both the subjects (railway infrastructure and vehicles) with a single sensing system but also provide cost-effectiveness and reliability in addition to other benefits. Therefore, in comparison to the conventional operational monitoring techniques, OFS-based systems can be considered to be more suitable with the provision of long-term solutions for several types of maintenance and other miscellaneous monitoring for both the vehicle and infrastructure.
