Maulana, A Review of RAMS Analysis Application on Railway System 57

A Review of RAMS Analysis Application on Railway System

Angga Ahmad Maulana 1*, I Wayan Suweca 2
1,2
Mechanical Engineering Study Program, Faculty of Mechanical and Aerospace Engineering, Bandung Institute
of Technology
Email: 1,*[email protected], [email protected]

DOI: https://doi.org/10.31284/j.jtm.2023.v4i2.4523
Received 27 May 2023; Received in revised 30 May 2023; Accepted 5 June 2023; Available
online 26 July 2023
Copyright: ©2023 Angga Ahmad Maulana, I Wayan Suweca
License URL: https://creativecommons.org/licenses/by-sa/4.0

Abstract
Railway systems play a crucial role in transportation infrastructure, requiring high levels of
reliability, availability, maintainability, and safety (RAMS) to ensure efficient and safe operations.
This contribution is an attempt to provide an overview of past and current research on the
application of RAMS to railway systems which, is broadly summarized into three review groups;
(1) review of RAMS applications on railway networks, (2) review of RAMS applications on railway
rolling stock, and (3) review of RAMS applications on railway infrastructure. Based on the review
results, the use of probabilistic criteria is widely used to calculate reliability, availability is
calculated using maintenance parameters of the system, while maintainability is calculated based
on reliability parameters and failure rates. For safety, it is obtained based on the characteristics of
the research object of each study. In the end, it is expected that methods from previous studies can
be used as a reference for the application of RAMS to the railway system in Indonesia.

Keywords: Availability, Maintainability, Railway System, Reliability, Safety

1. Introduction
The railway system can be described as a cohesive and integrated system, which encompasses
not only the physical infrastructure and rolling stock, but also the human resources and the various
norms, criteria, requirements, and procedures that are essential for the successful implementation of
railway transportation [1]. This system has existed since the 19th century and continues to develop
today. In the railway system, trains can be used for various purposes, such as for the transportation of
passengers, and goods, or even for military purposes where rails are used as the path of movement.
The railway system has several advantages, including being able to transport many passengers or
goods at once, being efficient in fuel use, and reaching a fairly high speed. In addition, it can also
reduce congestion on roads and help reduce air pollution. However, as with other transportation
systems, the railway system also has some disadvantages, such as high investment costs, intensive
maintenance needs, and limited available lines. With these weaknesses, an efficient and effective
railway management system is needed to improve the railway system's performance, both from a
technical and non-technical perspective.
The implementation of RAMS (Reliability, Availability, Maintainability, and Safety) for
railway management systems is crucial. In 1999, the European Union implemented the EN 50126
standard which aimed to restrict railway accidents and promote the security of railway activities for
the entirety of the system's life cycle [2]. The concept of RAMS comprehends four principal aspects,
specifically those of reliability, availability, maintainability, and system safety. In this regard, the
reliability of the railway system ensures that it operates according to expected specifications and
performance, the availability of the system ensures that the railway system is always available and can
be used to the best of its ability, the maintainability of the system ensures that the railway system is

*
Angga Ahmad Maulana ISSN: 2721-1878 | DOI: 10.31284/j.jtm.2023.v4i2.4523
58 Jurnal Teknologi dan Manajemen, Vol 4, No 2, Juli 2023: 57-77

easily repaired and routinely maintained, and the system's safety ensures that the railway system is
safe for all users [3].
Many railway organizations have implemented RAMS in their business activities. This is in
line with numerous studies and implementations related to the application of RAMS to railway
systems, which consist of various types of reviews, ranging from reliability only, reliability and
availability, and even those that review the overall aspects of RAMS. All of these are adjusted to the
needs of a railway organization, or the occurrence of specific problems related to railway assets that
require certain specific studies. For example, from the infrastructure aspect of railway tracks, research
has been conducted on the reliability of railway track maintenance operations [4], availability
approaches for railway track renewal operations operations [5], and reliability based on availability
factors for railway joints and crossings [6]. Regarding reliability, availability, and maintainability
(RAM), [2] has undertaken a comprehensive study on the track geometry as well as the various
railway track components. Research on signaling aspects are related to signaling safety assessment [7],
availability analysis of railway track circuits [8], and how to improve the reliability and availability of
track switching [9]. Meanwhile, the safety and availability evaluation of railway signaling systems has
been studied by [10]. From the rolling stocks aspect, the reliability and availability of locomotive
bogies have been studied by [11], while the reliability of freight car bogies has been studied using the
FMECA method by [12]. Then, [13]-[14] studied the reliability, availability, and maintainability of
diesel locomotives. Furthermore, from the operational aspect of railway systems, a railway network
model has been developed using RAMS analysis to improve the safety and stability of railway
network operations [15]. The balance between network availability for railway operations and track
construction activities has also been studied by [16]. Analytical models for reliability assessment have
also been conducted by [17] to predict the reliability and unreliability of railway subsystems. On the
other hand, RAMS development has been conducted by [18] by adding additional parameters that are
specifically related to railways. Therefore, if RAMS studies are applied properly to management, they
will result in a reliable railway system, effective and efficient maintenance, and optimal asset life cycle
cost (LCC).
The relationship between RAMS and LCC is very close because RAMS directly affects the
total lifecycle cost (LCC) of a system. The reliability, availability, and maintainability of a system
directly influence maintenance costs, component replacement costs, and repair costs, all of which are
important factors in calculating the LCC of a system. The integration of RAMS and LCC approaches
can help determine the Key Performance Indicators (KPIs) of a system so that the system's
performance can be known and used as a consideration in decision-making throughout the system's
lifecycle [19].
The purpose of this literature review is to review, summarize, and compile research that has
been conducted on RAMS analysis in railway systems. The railway system discussed in this paper
relates to infrastructure, rolling stocks, and railway network operations. This review is expected to
serve as a guide and reference for further application and development in railway asset management,
especially for railways in Indonesia. The systematic review will be presented as follows: (1) RAMS
application in railway network operation, (2) RAMS application in railway rolling stock, and (3)
RAMS application in railway infrastructure.

2. Methodology
This research was conducted by exploring existing journals as references by comparing the
methods and results of each journal. Each journal is analyzed related to the RAMS method, starting
from reliability, availability, maintainability, and safety, then compared with the results obtained and
the effectiveness of the method used for application in a case study. After getting methods from each
aspect of RAMS, then used as a reference in choosing which methods can be applied in the future to
be used on railways in Indonesia.

3. RAMS Review on Railway Systems

 Maulana, A Review of RAMS Analysis Application on Railway System 59

In this chapter, the author discusses the application of RAMS for each railway system review
into sub-chapters, starting from network, rolling stock, and railway infrastructure. Each sub-chapter
will be presented into each aspect of RAMS, starting from reliability, availability, maintainability, and
safety to facilitate categorization.

3.1. Review RAMS Application in Railway Networks

The railway network is currently endeavoring to tackle the challenges that have arisen due to
the increasing demands for capacity, speed, and mobility for the transportation of products and
passengers. This complex system comprises various technologies and a diverse range of stakeholders
[20]. The railway system is undergoing constant construction and development, resulting in a more
vital temporal and spatial dynamic between the network and the organization of rail lines. However,
the rapid growth has made operating and maintaining the entire railway network increasingly
challenging. Additionally, the trains traveling between stations have created more complex
relationships. In the event of a failure at a crucial station, the transportation efficiency of the entire
network would be reduced [21]. Therefore, with RAMS analysis of the railway network, it will help
railway organizations in determining the appropriate investments required to enhance safety
management in each section and station of the railway network [15]. Doing so could increase the
safety and resilience of the railway network's operation. In this section, the application of RAMS to
the railway network operation that has been done previously by several researchers will be discussed
which can be seen in Table 1.
Table 1. Journal Details and Aspects of RAMS Used for Railway Network
No Study Title Year Reliability Availability Maintainability Safety
Jack Development of an Extended
1 Litherland RAMS Framework for 2019 ✓ ✓ ✓ ✓
et al. [18] Railway Networks
RAMS analysis of railway
Zhe Zhang
2
et al. [15]
network: model development 2020 ✓ ✓ ✓ ✓
and a case study in China

Reference [18] proposed that assessment framework has the potential to be leveraged by
railway asset managers in order to allocate resources in a strategic manner to improve the railway
network. The augmented RAMS framework encompasses a hierarchical structure consisting of four
levels, which includes not only the conventional RAMS parameters but also supplementary parameters
such as security, health, environment, economics, and politics (SHEeP). The methodology is being
trialed on the TransPennine route, and if successful, it will be rolled out to other parts of the network.
In reference [15], the focus was on a RAMS analysis method for operating railway networks, which
takes into account the probability of failure for stations and sections, the anticipated loss of network
availability resulting from such failures, and a safety index for assessing the safety of railway stations
and sections. The proposed method is demonstrated through a real-world case study, revealing its
capability to identify high-risk sections and stations and determine appropriate investments in safety
management for each of them. Additionally, this study introduces a data-driven approach to examine
network reliability, develops a safety index, and suggests potential applications of the proposed model
to enhance the safety and resilience of railway network operations.

3.1.1. Reliability of Railway Networks

Reliability analysis by [18] is defined as the reliability of a component that can be calculated
mathematically using the failure / hazard rate, λ(𝑡). It can be seen that the failure rate is related to
reliability according to the following equation:
𝑡
−∫λ(τ)𝑑τ
𝑅(𝑡) = 𝑒 0
(1)
Reliability may be determined through the calculation of a constant failure rate as
−λ𝑡
𝑅(𝑡) = 𝑒 (2)
60 Jurnal Teknologi dan Manajemen, Vol 4, No 2, Juli 2023: 57-77

However, if the failure rate is not uniform, it becomes necessary to employ statistical models, such as
Weibull, Gamma, or Log-normal, to calculate reliability. In instances where the component can be
repaired, the Mean time between failure (MTBF) is utilized and expressed in the following manner.
𝑇𝑜𝑡𝑎𝑙 𝑇𝑖𝑚𝑒−𝑇𝑜𝑡𝑎𝑙 𝐷𝑜𝑤𝑛 𝑇𝑖𝑚𝑒
𝑀𝑇𝐵𝐹 = 𝑛
(3)
where, n is the failures number. The railway network consists of many components. Therefore
reliability analysis needs to be calculated as a system that consists of series and parallel systems. The
system reliability for n components in a series can be calculated based on the assumption of
component independence as
𝑛
𝑠
𝑅𝑠𝑦𝑠(𝑡) = ∏ 𝑅𝑖(𝑡) (4)
𝑖=1

For 𝑛 components in a parallel system, the system reliability can be expressed as follows:
𝑛
𝑠
𝑅𝑠𝑦𝑠(𝑡) = 1 − ∏ (1 − 𝑅𝑖(𝑡)) (5)
𝑖=1

In the study [18], the number of service affecting failures (SAFs) that occur is the basis for
calculating the reliability of the rail network, where SAFs are assumed to be: (1) independent, (2)
cause complete failure of the system.
Assuming these two premises, the railway system can be represented as a series of
interconnected assets without redundancy, and its reliability can be computed through standard
methods outlined in (4). Given that the system consists of numerous components arranged in series,
attaining a high level of reliability for a railway network is challenging.
Whereas, in [15], the physical network of the railway system can be characterized as G = (N,
E), wherein the node n ϵ N signifies the railway station while the link e ϵ E represents the railway lines
that connect these stations. Assuming that the two adjacent railway stations are denoted by i, j, the
railway line linking stations i and j can be represented by (i, j). The reliability of a rail network
represents the probability of accident-free operation. Accident reports allow us to determine the
𝑢𝑝 𝑑𝑜𝑤𝑛
number of accidents. Accidents can occur on both the up and down lines. Let 𝑃𝑖𝑗 and 𝑃𝑖𝑗 denote the
accident probabilities on the up and down lines connecting neighboring stations i and j, respectively,
and can be elucidated as follows.
𝑢𝑝 𝑑𝑜𝑤𝑛
𝑢𝑝 𝑁𝑖𝑗 𝑑𝑜𝑤𝑛 𝑁𝑖𝑗
𝑃𝑖𝑗 = 𝑇
, 𝑃𝑖𝑗 = 𝑇
(6)
𝑢𝑝 𝑑𝑜𝑤𝑛
where 𝑁𝑖𝑗 and 𝑁𝑖𝑗 represent the respective counts of accidents on the upward and downward lines,
while T denotes the statistical time period.
Hence, the probability of failure for the railway segment (i, j) can be expressed as follows:
𝑢𝑝 𝑑𝑜𝑤𝑛
𝑃𝑖𝑗 = 𝑃𝑖𝑗 𝑃𝑖𝑗 (7)
𝑘
There are many tracks at the train station for the trains to run on. Let 𝑃𝑖 be the accident
probability of line k at station i. As a result, the railway station's failure probability may be expressed
as:
𝐾
𝑘
𝑃𝑖 = ∏ 𝑃𝑖 (8)
𝑘=1

The railway network G’s reliability can be expressed as:

𝑅𝐺 = ∏ (1 − 𝑃𝑖𝑗) ∏ 1 − 𝑃𝑖
𝑖,𝑗∈𝑁 𝑖∈𝑁
( ) (9)
 Maulana, A Review of RAMS Analysis Application on Railway System 61

3.1.2. Availability of Railway Networks

Reference [18] has explained that the calculation of maintainability has not been accounted for
in the study. This implies that the use of MTTF and MTTR is not feasible. However, an alternative
approach to determine the availability is to consider the percentage of time that the network is unable
to run full service. Hence, the calculation of availability can be derived by aggregating the downtime
for all assets.
Whereas, reference [15] explained that from the viewpoint of a complex network, it is possible
to examine the accessibility of railway network G. The availability of the railway network G is
represented in this study by the network efficiency E and capacity C. Given the volume of traffic
between stations, the G's railway network efficiency may be summarized as follows
𝑢𝑚𝑛
𝐸= ∑ 𝑑𝑚𝑛
(10)
𝑚𝑛

where 𝑢𝑚𝑛 represents the number of trains commuting from station m to station n and 𝑑𝑚𝑛 signifies the
least distance between station m and station n, E' may be utilized to denote the efficiency of the
network in the event of a station or section malfunction. The deterioration of network efficiency can be
explicated as follows:
'
𝐸−𝐸
∆𝐸 = 𝐸
(11)
Then, the trains maximum number that may be put into the network is referred to as the
network capacity. Because it can assess both directional and unidirectional networks, the I-O approach
is utilized to gauge the railway network's capacity. Let 𝑥𝑖𝑗 represent the trains number between the
nearby railway stations of i and j. Let 𝑋𝑗 stand for the number of trains arriving at station j from other
stations. The following is a description of the I-O matrix:
𝑖
𝑖 𝑥𝑗
𝐹𝑗 = 𝑋𝑗
(12)

The trains number arriving at station j as follows:

𝑁−1 𝑆 𝑁−1
𝑖 𝑠 𝑖
𝑋𝑗 = ∑ 𝐹𝑗𝑋𝑖 + ∑ 𝑥𝑗 = ∑ 𝐹𝑗𝑋𝑖 + 𝑈𝑗 (13)
𝑖=1 𝑠=1 𝑖=1
𝑠
where 𝑥𝑗 is utilized to represent the count of trains that originate from stations situated outside of the
railway network and terminate at station j. (13) can be defined into matrix form as follows:
(−𝑖)
𝑋 =𝐹 𝑋+𝑈 (14)
where the trains' total number departing from network stations is represented by U, whereas the trains'
total number arriving at all network stations is represented by X. The network capacity can be defined
as follows:
(−𝑛) −1
𝑋 = (1 − 𝐹 ) 𝑈 (15)
If the train service along the stretch connecting stations u and v becomes inoperable, it would result in
a modification of the network's capacity as:
−(𝑢,𝑣) (−𝑖−(𝑢,𝑣)) −1
𝑋 = (1 − 𝐹 ) 𝑈 (16)
Therefore, network capacity loss ∆𝐶 can be expressed as follows:
−(𝑢,𝑣)
∑𝑋−∑𝑋
∆𝐶 = (17)
∑𝑋
62 Jurnal Teknologi dan Manajemen, Vol 4, No 2, Juli 2023: 57-77

3.1.3. Maintainability of Railway Networks

How quickly a system can be fixed after a failure and how much downtime ensues from the
breakdown are measured by maintainability. The less complicated the system is to repair, the quicker
the repairs may be completed and the shorter the downtime will be. On the railway network system,
[18] proposed a method to calculate maintainability with the following equation:
−𝑡

𝑀(𝑡) = 1 − 𝑒 𝑀𝑇𝑇𝑅 (18)

where 𝑀(𝑡) represents the probability that the specific component shall be rectified within the
designated time frame t, MTTR is the mean time to repair. The maintainability in this study will be
calculated as the number of the hours spent on each of the planned and unplanned maintenance
activities in Figure 1.
Furthermore, reference [15] explained that the accident reports have the recuperation time
noted. Let ts and te stand for the accident's beginning and ending times. Maintenance M is defined as
the ratio of the normal operation time to the total operation time of the railway network., which is
expressed as follows
𝑡𝑟
𝑀=1− 𝑡0
(19)

where tr = te – ts and to represents the duration of operation for each day.

Figure 1. Maintenance Actions During Unplanned and Planned Processes

3.1.4. Safety of Railway Networks

Reference [18] explained that it is challenging to define a measure since there is no accepted
concept of safety on the railway. The safety metric has to include a variety of direct and indirect safety
components. Fatalities and Weighted Injuries (FWI) are a typical indicator of safety performance on
UK railways. The "Safety Risk Model" and "Precursor Indicator Model" were created by NR and the
Rail Safety and Standards Board (RSSB) to determine the likelihood of the "Top Event" and the
 Maulana, A Review of RAMS Analysis Application on Railway System 63

consequent number of injuries and fatalities. In this study, these models will be utilized to evaluate the
network's performance in terms of safety.
However, reference [15] explained that the possibility of an accident, the loss of availability,
and maintenance (M) should all be taken into account when evaluating the safety or risk of the railway
network. The accident consequence S, which may be used to gauge how safe a railway network is, is
as follows:
𝑆 = (1 − 𝑅)∆Α(1 − 𝑀) (20)
where, S signifies the outcome of an accident, (1 – R) represents the likelihood of an accident
occurring within the railway network, ∆Α signifies the magnitude of unavailability that would ensue in
the absence of maintenance, (1 – M) denotes the impact of maintenance on ensuring operational safety.
The network-scale safety, however, cannot account for the significance of each station's or section's
safety. Consequently, equation (20) is derived based on the specific categories of section failure and
station failure. When section (i, j) is disrupted due to operational accidents, the associated risk can be
described in the following manner:
𝐸𝐿𝑖𝑗 = 𝑝𝑖𝑗∆𝐸(1 − 𝑀), 𝐶𝐿𝑖𝑗 = 𝑝𝑖𝑗∆𝐶(1 − 𝑀) (21)
where 𝐸𝐿𝑖𝑗 denotes the expected loss of network efficiency and 𝐶𝐿𝑖𝑗 denotes the expected loss of
network capacity. If the station i is broken by the operation accidents, the risk can be described as
follows:
𝐸𝐿𝑖 = 𝑝𝑖𝑗∆𝐸(1 − 𝑀), 𝐶𝐿𝑖 = 𝑝𝑖𝑗∆𝐶(1 − 𝑀) (22)
where 𝐸𝐿𝑖 denotes the expected loss of network efficiency and 𝐶𝐿𝑖 denotes the expected loss of
network capacity.
Evaluating the safety of a section using (21) and (22) can yield varying results. To ensure a
comprehensive evaluation, it is necessary to consider the loss of network efficiency and capacity when
assessing the level of safety or each section or station grade. To accomplish this, the K-means method
was utilized, a clustering method capable of classifying data into distinct groups, to identify critical
sections and stations with regard to safety.

3.1.5. Summary of RAMS Application in Railway Networks

The review in this section obtained the following results which can be seen in the Table 2.
Table 2. Assessment Method for RAMS Used for Railway Network in Included Studies
No Study Reliability Availability Maintainability Safety
Fatalities and
Jack Litherland et al. MTBF,
1 MTTF/(MTTF+MTTR) MTTR Weighted Injuries
[18] Statistical model
(FWI)
Possibility of an
Proportion of
Probabilistic Network efficiency E, accident, loss of
2 Zhe Zhang et al. [15] normal operation
Criteria Network capacity C availability and
time
maintenance

The findings indicate that the approaches employed in each examination for every facet of
RAMS are not notably identical, given that every utilized technique is best suited for the particular
research target of the respective study. In particular, Reference [18] employs a greater number of
maintenance parameters to derive reliability, availability, and maintainability values, whereas [15]
employs time and accident parameters to derive reliability, maintainability, and safety, and availability
utilizes the capacity and efficiency parameters of a network.

3.2. Review RAMS Application in Railway Rolling Stocks

Railway rolling stocks are vehicles that can move on the railway track [1]. It is composed of
multiple subsystems, including the car body, bogie, braking system, propulsion, and power supply
[22]. A proficient and properly serviced rolling stock system is imperative for achieving the railway
64 Jurnal Teknologi dan Manajemen, Vol 4, No 2, Juli 2023: 57-77

system's objective of enhancing performance. Nevertheless, the intricate structures and numerous
components of the system pose a challenge to conducting an objective evaluation of its performance
[23]. Therefore, applying RAM to railway rolling stock will make maintenance activities effective and
optimize life cycle costs. In this section, the application of RAMS to rolling stock will be reviewed,
where the references reviewed can be found in Table 3.
Table 3. Journal Details and Aspects of RAMS Used for Railway Rolling Stock
No Study Title Year Reliability Availability Maintainability

Measurement and Evaluation of

D. Bose et Reliability, Availability and
1
al. [24] Maintainability of a Diesel
2013 ✓ ✓ ✓
Locomotive Engine
Maciej Assessment of Reliability
2 SzkodA Availability and Maintainability of 2014 ✓ ✓ ✓
[25] Rail gauge change system
Maciej Analysis of reliability, availability
3 SzkodA and maintainability (RAM) of 2014 ✓ ✓ ✓
[13] SM48 diesel locomotive

Reference [24] centers its attention on the reliability and availability facets of a noteworthy
constituent within a diesel locomotive engine for railway applications. The application of the Weibull
distribution was employed for reliability analysis, and an array of data plots and failure rate
information were utilized to attain outcomes that are instrumental in mitigating unforeseen failures
while enhancing the engine's reliability and availability. Then, reference [25] centers on conducting a
comparative evaluation of the dependability of two distinct rail gauge replacement systems, namely,
carriage bogie swapping and the SUW 2000 self-adjusting wheelset. This assessment is achieved by
considering dependability as a comprehensive characteristic encompassing reliability, availability, and
maintainability. The analysis results show that the SUW 2000 system has a higher failure rate than the
bogie wagon exchange system with the average failures number in operation one year (MNF) for the
SUW 2000 system more than twice as high as the bogie wagon exchange system. Meanwhile,
reference [13] centers on conducting RAM analysis on SM48 diesel locomotives, drawing on
operational tests administered on a particular subset of locomotives operated by the Polish railway
company PKP CARGO S.A.The outcomes of the RAM analysis serve as a foundation for altering the
maintenance cycle of the locomotives, while the reliability ratio established in the analysis can
function as a cornerstone for Life Cycle Cost (LCC) analysis.

3.2.1. Reliability of Railway Rolling Stock

In reference [24], the railway diesel locomotive engine is divided into two main subsystems,
that are compressor and vehicle and structures. In his study, the prediction of reliability, if the failure
rate is constant, is the same as that of reference [18] on the railway network, as well as in calculating
MTBF and MTTF.
Because the failure rate of locomotive components is not constant, the Weibull distribution model can
be employed. Furthermore, the method of linear regression analysis verifies the appropriateness of
utilizing the Weibull distribution for the different constituents of this study. The scrutiny establishes
the optimal-fit line in the context of the least square approach. The least square test was conducted to
ascertain the increasing/decreasing rate of failures.
The utilization of the Probability equation has facilitated the conduct of a linear regression analysis

(∑𝑥)[∑𝑓(𝑥)]
∑[𝑥𝑓(𝑥)]− 𝑁
𝑅(𝑥), 𝑓(𝑥) = (23)
2
2
(∑𝑥) ∑𝑓(𝑥)
2 2
[∑(𝑥 )− 𝑁
][∑𝑓(𝑥 )− 𝑁
]
 Maulana, A Review of RAMS Analysis Application on Railway System 65

where, 𝑥: The duration of breakdowns, 𝑓(𝑥): The cumulative percentage of failures, 𝑁: trials number,
𝑅(𝑥), 𝑓(𝑥): coefficient of correlation that can be seen in Table 4.
Table 4. Correlation Coefficient
Components’s name Coefficient of Correlation
Compressor 0.783
Vehicle and Stuctures 0.741

Weibull distribution can be expressed as follows:

𝑡 β
𝑡 β−1 − ( )
𝑓(𝑡) = ( )( )
β
η η
𝑒 η
(24)

where, η: life, β: Values of the shape factor. The shape of the distribution is determined by the beta (β)
value.

Figure 2. Weibull Distribution

The present examination has condensed its findings to the parameter of Weibull shape. This
parameter, which is denoted by β, likewise functions as an indicator of whether the rate of failure is
uniform, escalating, or declining. Specifically, when β equals 1.0, β exceeds 1.0, or β falls below 1.0,
the corresponding implication is that the rate of failure remains consistent, amplifies, or reduces,
respectively.
Reference [25] explained that calculating the reliability of elements, subsystems and systems
using relevant reliability ratios. To enable a comparative analysis of the reliability of different systems,
the MNF ratio was employed, which represents the average number of failures that occur during a year
of operation. This particular ratio is explicitly defined as follows when applied to a singular
component:
𝑀𝑁𝐹𝑖 = ( ) 𝐻𝑖(𝑡)
𝑇𝑖
. 8, 760. 0 ⎡
⎣
𝑓𝑎𝑖𝑙𝑢𝑟𝑒𝑠
𝑦𝑒𝑎𝑟
⎤
⎦
(25)

where, 𝑀𝑁𝐹𝑖: refers to the average number of failures experienced by element "i" during a year of
operation. The renewal function of element "i" in the maintenance cycle is represented by 𝐻𝑖(𝑡), while
𝑇𝑖: denotes the duration of operation of element "i" in the maintenance cycle, expressed in hours.
For a subsystem, the total failures of failures mean number during a single year of operation is:
𝑀𝑁𝐹𝑠 = 𝑀𝑁𝐹1 + 𝑀𝑁𝐹2 + … + 𝑀𝑁𝐹𝑛 = ( 𝐻1(𝑡)
𝑇1
+
𝐻2(𝑡)
𝑇2
+…+
𝐻𝑛(𝑡)
𝑇𝑛 )
. 8, 760 ⎡
⎣
𝑓𝑎𝑖𝑙𝑢𝑟𝑒𝑠
𝑦𝑒𝑎𝑟
⎤
⎦
(26)

In the aforementioned formulation, the renewal function denoted as 𝐻(𝑡) is utilized. This
function, under the assumption that the renewal period is considerably insignificant in comparison to
66 Jurnal Teknologi dan Manajemen, Vol 4, No 2, Juli 2023: 57-77

the temporal extent of appropriate functioning of the subject entity, articulates the projected quantity of
renewals that is equivalent to the failures number up until time t. Moreover, the renewal function is
defined as follows:
∞
𝐻(𝑡) = ∑ 𝐹𝑛(𝑡) (27)
𝑛=1

where, 𝐹𝑛(𝑡) is the distribution function characterizing the operation of the object until the n-th failure,
which is subject to renewal.
Reference [13] expounds upon the methodology of determining the reliability ratio through
the employment of the Weibull distribution. This statistical technique allows for the estimation of the
parameters of the probability density function concerning the duration of proper operation before
failure. The formula for said calculation is as follows:
𝑡 𝑎
𝑡 𝑎−1 − ( )
𝑓(𝑡) = ( )( )
𝑎
𝑏 𝑏
𝑒 𝑏
(28)
And the function that represents the cumulative distribution of time elapsed until system failure has
been derived:
𝑡 𝑎
𝐹(𝑡) = 1 − 𝑒𝑥𝑝⁡ − ( 𝑏 ) (29)
Meanwhile, the mean time to failure is computed using the appropriate operational distribution
function, expressed as:
∞
𝑀𝑇𝑇𝐹 = ∫ 𝑡. 𝑓(𝑡)𝑑𝑡 (30)
0

Furthermore, the mean time between failures for the locomotive is determined using the distribution
function of the time between failures, formulated as follows:
∞
𝑀𝑇𝐵𝐹 = ∫ 𝑡. 𝑓𝑘(𝑡)𝑑𝑡 (31)
0

3.2.2. Availability of Railway Rolling Stock

Reference [24] defines availability as the likelihood that the system is functioning
appropriately upon request for utilization which can be expressed in three types depending on
consideration of the time elements.

(1) Inherent Availability

This concept pertains to the likelihood of a system or equipment functioning adequately when
utilized within specified circumstances in an optimal support setting, without factoring in any
scheduled or preventative maintenance at any point in time. Its mathematical representation is as
follows:
𝑀𝑇𝐵𝐹
𝐴𝐼𝑁 = (𝑀𝑇𝐵𝐹+𝑀𝑇𝑇𝑅) (32)

(2) Achieved Availability

This is a measure of the likelihood that a specific system or equipment will function
effectively when utilized in prescribed conditions within an optimal support environment, at any given
moment. Such probability can be denoted as:
𝑀𝑇𝐵𝑀
𝐴𝑎 = (𝑀𝑇𝐵𝑀+𝑀𝐷𝑇) (33)
where, MTBM refers to the average duration of active maintenance downtime that arises from both
preventive and corrective maintenance procedures.

(3) Operational Availability

 Maulana, A Review of RAMS Analysis Application on Railway System 67

Operational availability is defined as the likelihood of a system or equipment functioning

acceptably in accordance with specified conditions and within a real supply context at any given
moment. This metric can be articulated as follows:
𝑀𝑇𝐵𝐹
𝐴𝑂𝑃 = (𝑀𝑇𝐵𝐹+𝑀𝐷𝑇) (34)
Furthermore, reference [25] assuming that each and every constituent element functioning
within a given system is depicted through probability distribution functions of the same nature, with
respect to both the operation duration and restoration, the system availability, denoted as A(t), can be
represented by the ensuing function
𝑡
𝐴(𝑡) = 1 − 𝐹(𝑡) + ∫[1 − 𝐹(𝑡 − τ)]ℎ(τ)𝑑τ (35)
0
𝐻(𝑡)
where, ℎ(𝑡): renewal density function : ℎ(𝑡) = 𝑑𝑡
The aforementioned equation is infrequently utilized in practical applications due to its
significant level of computational intricacy. Usually, the technical availability ratio, which represents
the average duration when the system in question is operational, is commonly used instead.
𝐴(∞) = 𝐴(𝑡) (36)
The availability ratio denotes the mean technical availability throughout the maintenance
cycle, particularly between successive maintenance (revision) events of the scrutinized entities. The
availability ratio of an individual entity is articulated as follows:
𝑇𝑍𝑖
𝐴𝑖 = 𝑇𝑍𝑖+𝑇𝑁𝑖+𝑇𝑂𝑖
(37)

where:
𝑇𝑍𝑖: average duration of availability for item "i" (in hours).
𝑇𝑁𝑖: average duration of unavailability for item "i" caused by corrective maintenance (in hours).
𝑇𝑂𝑖: average duration of unavailability for item "i" resulting from preventive maintenance activities
(in hours).
In reference [13], the ratio of operational availability Ao and the ratio of actual availability AR
were used. The operational availability ratio can be expressed as follows:
𝑇𝑍
𝐴𝑜 = 𝑇𝑍+𝑇𝑁 (38)
Meanwhile, the actual availability ratio was defined as:
𝑇𝑍
𝐴𝑅 = 𝑇𝑍+𝑇𝑁+𝑇𝑂 (39)

𝑇𝑂 denotes the mean duration of locomotive immobilization during scheduled downtimes and upkeep
operations amidst consecutive revisions, which can be mathematically represented as follows

𝑇𝑂 = ∑(𝑃𝑈𝑖. 𝑀𝑇𝑇𝑅𝑖) (40)

where:
𝑃𝑈𝑖: count of scheduled downtimes or activities of maintenance
𝑀𝑇𝑇𝑅𝑖: average duration of scheduled downtimes or maintenance activities.

3.2.3. Maintainability of Railway Rolling Stock

Reference [24] establishes a correlation between maintainability and maintenance strategies,
based on which several novel strategies have been deployed as maintenance strategies aimed at
mitigating equipment breakdown-related concerns. A few of the widely used maintenance strategies
include:
68 Jurnal Teknologi dan Manajemen, Vol 4, No 2, Juli 2023: 57-77

Table 5. The Strategy of Maintenance

Maintenance Strategy Maintenance Approach Signification
Corrective Fix it when broke Large maintenance budget.
Preventive Scheduled Maintenance Peplacement of periodic component
Maintenance decision based on
Predictive Condition based Monitoring
equipment condition.
Failure correcting root causes and
Proactive Failures sources detection
monitoring

Reference [25] explained the utilization of mean maintenance time (MMT) in a year of
operation. This variable encompasses the entirety of time allocated towards corrective and preventive
maintenance of a system and is utilized to compare system maintainability. This ratio is defined as
follows for one element:
(𝐻𝑖(𝑡).𝑀𝑇𝑇𝑅𝐵𝑖)+(𝑁𝑃𝑀𝐴𝑃𝑖.𝑀𝑇𝑇𝑀𝑃𝑖)+(𝑁𝑃𝑀𝐴𝑁𝑖.𝑀𝑇𝑇𝑀𝑁𝑖)
𝑀𝑀𝑇𝑖 = ( 𝑇𝑖 ) ℎ𝑟𝑠
. 8, 760. 0 ⎡ 𝑦𝑒𝑎𝑟 ⎤
⎣ ⎦
(41)

where:
𝐻𝑖(𝑡): function of renewing element "i" within the maintenance cycle.
𝑀𝑇𝑇𝑅𝐵𝑖: average duration required to restore element "i" (in hours).
𝑁𝑃𝑀𝐴𝑃𝑖: quantity of scheduled maintenance activities performed on element "i" within the
maintenance cycle.
𝑀𝑇𝑇𝑀𝑃𝑖: average duration required to conduct periodic inspections for element "i" (in hours).
𝑁𝑃𝑀𝐴𝑁𝑖: quantity of maintenance activities for revising element "i" within the maintenance cycle.
𝑀𝑇𝑇𝑀𝑁𝑖: average duration required to perform maintenance revisions on element "i" (in hours).
𝑇𝑖: duration of element "i" being in operation within the maintenance cycle (in hours).
The total average maintenance time for a subsystem during a year of operation is:
ℎ𝑟𝑠
𝑀𝑀𝑇𝑠 = 𝑀𝑀𝑇1 + 𝑀𝑀𝑇2 + … + 𝑀𝑀𝑇𝑛 ⎡ 𝑦𝑒𝑎𝑟 ⎤ (42)
⎣ ⎦
Meanwhile, reference [13] explain that the evaluation of the maintainability of the SM48
locomotive encompasses both planned and preventative maintenance efforts conducted throughout its
maintenance regimen. The distribution function G(t) for renewal time, both empirical and theoretical,
as well as the average time required for repair (MTTR), has been established about scheduled
maintenance. The MTTR encompasses both repair time and the technical delays incurred during
diagnostic procedures and the procurement of spare parts. The locomotive renewal time cumulative
distribution function as follows:

(
𝐺(𝑡) = 0. 5 1 + Φ ( 𝐿𝑛(𝑡)−𝑎
δ 2 )). 𝑓𝑜𝑟 𝑡 > 0 (43)

where:
𝑧
2
Φ(𝑧) =
2
Π
∫ exp 𝑒𝑥𝑝 − 𝑡 ( ) 𝑑𝑡 🡪 Gauss distribution function (44)
0

The renewal time distribution function has been utilized to determine the mean renewal time, as
follows:
( )
2
δ
𝑀𝑇𝑇𝑅𝐵 = 𝑒𝑥𝑝 𝑎 + 2
(45)

3.2.4. Summary of RAMS application in railway rolling stock

The review in this section obtained the following results which can be seen in the Table 6.
Table 6. Assessment Method for RAMS Used for Railway Rolling Stock in Included Studies
 Maulana, A Review of RAMS Analysis Application on Railway System 69

No Study Reliability Availability Maintainability

D. Bose et al. Common maintenance
1 Statistical model (Weibull) MTBF, MTTR, MTBM, MDT
[24] strategy
Probabilistic criteria,
Maciej SzkodA Mean number of failures
2 𝑇𝑍𝑖 MTTR, MTTM
[25] (MNF) 𝐴𝑖 = 𝑇𝑍𝑖+𝑇𝑁𝑖+𝑇𝑂𝑖

Maciej SzkodA 𝑇𝑍 Distribution function,

3 Statistical model (Weibull) 𝐴𝑅 =
[13] 𝑇𝑍+𝑇𝑁+𝑇𝑂 MTTR

The findings indicate that the Weibull distribution is extensively employed for reliability
computation owing to the non-constant failure rate of railway systems, necessitating the adoption of an
appropriate approach for obtaining the reliability value. Moreover, nearly all maintenance criteria are
utilized for achieving availability and maintainability as a consequence of reliability.

3.3. Review RAMS Application in Railway Infrastructure

Railway infrastructure generally consists of railways track, railway stations, and railway
operating facilities [1]. In this section, a review of the application of RAMS to railway infrastructure
will be discussed from the research that has been conducted. The RAMS journals of railway
infrastructure reviewed by the author will be more related to railways track because many previous
studies have discussed these topics. Not all the research journals in apply all aspects of RAMS to their
analysis. Therefore, the authors have divided each study according to the reviewed aspects and the
details of the research journals can be seen in Table 7.
Reference [26] focused on implementation of the EN 50126 standard and the use of RAMS
analysis techniques can improve the safety and reliability of railway systems. The investigation carried
out on the railway system in Uzbekistan demonstrates that the implementation of RAMS analysis can
be effectively employed to discern potential hazards and enhance the efficacy of maintenance and
restoration endeavors on railway systems. Reference [19] focused on rail performance assessed by
incorporating life cycle cost (LCC) and availability, maintainability, and safety (RAMS) approaches
into key performance indicators (KPIs). These KPIs can quantitatively influence decision-making
throughout the railway lifecycle (design, maintenance, and renewal). Reference [2] has directed its
focus toward the application of Reliability, Availability, and Maintainability (RAM) analysis in the
context of railway infrastructure management. The study incorporates a case study of the high-speed
ballasted railway line between Tashkent and Sirdaryo railway stations in Uzbekistan. The study's
objective is to identify the individual reliability level of infrastructure parameters and system
reliability level for IL (Intervention limit), and IAL (Immediate action limit) limits. The study's results
indicate that the individual reliability level of infrastructure parameters ranges from 62% to 89% for IL
and IAL limits. However, the system reliability level is found to be between 13% and 68% for IL and
IAL limits, respectively. This implies that any issue arising with one of the infrastructure parameters
necessitates a restriction or closure of railway line operations. This study shows that periodic planned
and corrective maintenance methods are preferable for Uzbekistan's railways. Reference [27]
conducted a comprehensive analysis of reliability, availability, maintainability, and safety (RAMS)
applications in the railway industry. Subsequently, a novel framework referred to as extended RAMS
(ExRAMS) was proposed for railway infrastructure, which integrates diverse RAMS analysis
approaches employed in the railway sector into a unified approach. The ExRAMS framework
comprises ten parameters arranged in a four-level hierarchy for evaluating RAMS performance. The
framework facilitates railway asset managers in assessing the attributes and current state of the railway
infrastructure, comparing different network components, and evaluating the requirements of various
stakeholders.
Table 7. Journal Details and Aspects of RAMS Used for Railway Infrastructure
No Study Title Year Reliability Availability Maintainability Safety
Hidirov and Railway Infrastructure
1
Guler [26] Reliability, Availability,
2017 ✓ ✓ ✓
70 Jurnal Teknologi dan Manajemen, Vol 4, No 2, Juli 2023: 57-77

Maintainability and Safety

(RAMS) analysis
Proposal of a Key
Praticò and Performance Indicator for
2
Giunta [19] Railway Track Based on
2018 ✓ ✓ ✓ ✓
LCC and RAMS Analyses
Reliability, availability,
Hidirov and and maintainability
3
Guler [2] analyses for railway
2019 ✓ ✓ ✓
infrastructure management
An alternative approach
to railway asset
Litherland
4 et al. [27]
management value 2021 ✓ ✓ ✓ ✓
analysis Application to
a UK railway corridor

3.3.1. Reliability of Railway Infrastructure

From the findings in this journal review, almost all studies discuss reliability. References [2],
[19], [26], and [27] have examined the track system side. The study conducted by [26] on reliability
commences with the assessment of the railroad track, wherein each track component is scrutinized
based on predetermined threshold values. Subsequently, the evaluation is conducted at the levels of the
analysis segment and maintenance segment. The statistical analysis of the data gathered from the
measuring instruments is performed, and this includes the determination of the type of reliability
function and its coefficients. The approach adopted is akin to the computation of the railway network
and rolling stock's reliability probability, as demonstrated in references [18] and [24], which employ
the failure rate (λ) as a parameter.
The work of [19] utilizes reliability measures that are expressed through the mean time
between failures (MTBF) and mean distance between failures (MDBF), both of which are indicative of
failure rates. It has been observed that the values of MTBF and MDBF tend to decrease as the traffic
load increases. Concerning MTBF and MDBF, the outcomes indicate:
𝑀𝐺𝑇
− τ
1𝑛
𝑎1−𝑏1.𝑒
𝑀𝑇𝐵𝐹 = 1 − −
𝑀𝐺𝑇
τ
(46)
1𝑑
𝑐1−𝑑1.𝑒
𝑀𝐺𝑇
− τ
2𝑛
𝑎2−𝑏2.𝑒
𝑀𝐷𝐵𝐹 = 1 − −
𝑀𝐺𝑇 (47)
τ
2𝑑
𝑐2−𝑑2.𝑒

where MGT = traffic expressed in millions of gross tons, 𝑎1 = 𝑏1 = 𝑐1 and 𝑎2 = 𝑏2 = 𝑐2, while
τ𝑖𝑛, τ𝑖𝑑, 𝑎𝑖, 𝑏𝑖, 𝑐 , and 𝑑𝑖 are the calibration factor coefficients.
𝑖
𝑀𝐺𝑇 = 𝑀𝐺𝑇𝑅 + 𝑀𝐺𝑇𝑇. α (48)
where 𝑀𝐺𝑇𝑅 = total traffic accumulated since the last restoration, 𝑀𝐺𝑇𝑇 = Total accumulated traffic
resulting from the construction, and α = coefficient that accounts for minor effects. From (46) and
(47), reliability is defined as
𝑅 = 𝑀𝑇𝐵𝐹. 𝑀𝐷𝐵𝐹 (49)
For the reliability analysis of the study conducted by [2], several statistical programs were
employed to fit the distribution to both track geometry and track data components. The fit test
methods, including Chi-square, Kolmogorov-Smirnov, Anderson-Darling, and Shapiro-Wilk, were
utilized to derive a distribution function that conforms to the collected data set. Concerning track
geometry, it was discovered that the Gumbel Max distribution is exceptionally well-suited to the
twisting data set, while the four-parameter Burr (4P) distribution is appropriate for the gauge data set,
and the three-parameter Dagum (3P) distribution proves suitable for the alignment, cant, and leveling
 Maulana, A Review of RAMS Analysis Application on Railway System 71

data sets. The Gumbel Max (Maximum Extreme Value Type 1) probability density function (pdf) is
provided as follows:
1
𝑓(𝑥) = σ exp 𝑒𝑥𝑝 (− 𝑧 − exp 𝑒𝑥𝑝 (− 𝑧) ) (50)
where 𝑥 is independent variable, σ is continuous scale parameter (σ > 0) and µ is continuous location
𝑥−μ
parameter. 𝑥 is between − ∞ < 𝑥 <+ ∞ and 𝑧 = σ . The Gumbel Max (Maximum Extreme Value
Type 1) cumulative distribution function (cdf) is given by
𝐹(𝑥) = exp 𝑒𝑥𝑝 (− exp 𝑒𝑥𝑝 (− 𝑧) ) (51)
The provided equation represents the probability density function (pdf) of the three-parameter
Dagum distribution:
𝑥 α𝑘−1

𝑓(𝑥) =
( )
α𝑘 β
(52)
𝑘+1
β(1+( ) )
𝑥 α
β

where 𝑥 is independent variable, 𝑘 is continuous shape parameter (𝑘 > 0), α is continuous shape
parameter (α > 0) and β is continuous scale parameter (β > 0). The three-parameter Dagum
cumulative distribution function (cdf) is given by
−𝑘

(
𝐹(𝑥) = 1 +
𝑥 −α
( )
β ) (53)

The four parameters Burr density function (pdf) is given by

𝑥−γ α−1

𝑓(𝑥) =
( )
α𝑘 β
(54)
𝑘+1
β(1+( )) 𝑥−γ α
β

where 𝑥 is independent variable, 𝑘 is continuous shape parameter (𝑘 > 0), α is continuous shape
parameter (α > 0) and β is continuous scale parameter (β > 0) and γ is continuous location
parameter (γ = 0 yields the three-parameter Burr distribution). The four-parameters Burr cumulative
density function (cdf) is given by
−𝑘
𝐹(𝑥) = 1 − 1 + ( ( ))
𝑥−γ α
β
(55)

Based on the study results, the cumulative density function is a factor considered to determine
the reliability of track geometry parameters. The resulting reliability function can be used to calculate
the reliability level of each track geometry parameter following the established railway organization
limits.
For track components, reference [2] conducted a study of rails, sleepers, and ballast layers.
The measurement and damage criteria for track components refer to Uzbekistan Railways regulations,
namely for rail damage head-loss and gauge face wear has a maximum limit of 18 mm. The sleepers
are defined into three damages: damaged sleepers, cracked sleepers, and broken sleepers, with
cracking criteria below 0.2 mm, between 0.2 and 1.0 mm, and above 1.0 mm. As for the ballast layer,
the damage criteria are defined as the fouling characteristics and strain-shear modulus of clean and
dirty ballast. Such ballast fouling can cause degradation of the ballast layer and ballast elastic
modulus.
The three components were evaluated using the Johnson SB distribution method for reliability
analysis. The Johnson SB distribution is considered to be a better fit for the damage data obtained
compared to other statistical methods. The mathematical definition of the Johnson SB (Maximum
Extreme Value Type 1) distribution function for the probability density function (pdf) is given by
𝑓(𝑥) =
δ
λ 2π𝑧(1−𝑧)
𝑒𝑥𝑝 −( 1
2 (γ + δ ln 𝑙𝑛 ( ) ) )
𝑧
1−𝑧
2
(56)

where 𝑥 is the variable of independent, γ is continuous shape parameter, δ is continuous shape

parameter (δ > 0), λ is continuous scale parameter (λ > 0) and ζ is continuous location parameter. 𝑥
72 Jurnal Teknologi dan Manajemen, Vol 4, No 2, Juli 2023: 57-77

𝑥−ζ
is between ζ≤𝑥≤ζ + and 𝑧 = λ
. The Johnson SB (Maximum Extreme Value Type 1) cumulative
distribution function (cdf) is given by
((
𝐹(𝑥) = Φ γ + δ ln 𝑙𝑛 1−𝑧
𝑧
( ) )) (57)
where the Laplace Integral, denoted as Φ, is utilized in determining the cumulative density function,
which is a crucial factor in assessing the reliability of track component parameters.
From the reliability of track components and track geometry, the system's overall reliability,
either in series or in parallel, is calculated. The overall reliability for a series system is defined as
follows:
𝑛
𝑅𝑖 = ∏ 𝑟𝑖𝑗 (58)
𝑗

The overall reliability for parallel systems is defined as follows:

𝑛
𝑅𝑖 = 1 − ∏ (1 − 𝑟𝑖𝑗) (59)
𝑗
𝑡ℎ 𝑡ℎ 𝑡ℎ
where 𝑅𝑖: 𝑖 section overall reliability, 𝑟𝑖𝑗: 𝑗 component reliability at 𝑖 section, 𝑛: component
number.
The mathematical definition of reliability analysis, as presented by [27], is the component or system
probability to continue functioning from its initial time to a certain time τ, provided that it was time
zero operational. This probability can be determined using formula:
∞
𝑅(𝑡) = ∫ 𝑓(𝑥)𝑑𝑥 (60)
𝑡

where f (x) is the distribution of failure times of probability density function.

Due to the lack of independence among numerous components, the analytical determination of
𝑓(𝑥) is often impractical in railway systems. Moreover, numerous interdependent components are
arranged in series and parallel, making it impractical to calculate reliability using conventional
methods. As a result, railway asset managers resort to various metrics to describe their system's
reliability, none of which adhere to the mathematical definition presented in (60). Reference [27]
introduces two metrics that can evaluate the railway system reliability: number of service-affecting
failures (SAFs) and mean time between SAFs.

3.3.2. Availability of Railway Infrastructure

The demand for higher availability in railway operations is constantly increasing, and the
availability of railway systems and subsystems significantly impacts the overall operational
availability [28]. In railway systems, enhancing the availability of infrastructure is a crucial objective,
as accidents or incidents often result in substantial losses. Availability refers to a system's capacity to
execute a necessary function under specified conditions at a particular moment [29]. The application
of RAMS on railways related to availability will be discussed in this section.
Reference [26] formulates availability as the period when a line can be opened or closed for
operation. Below is the mathematical representation of the availability analysis:
𝑇𝑖
𝐴𝑖 = 𝑇𝑖+𝑡𝑖
(61)

where 𝐴𝑖 is the availability, 𝑇𝑖 is the total time (365 days), dan 𝑡𝑖 is the time the line is closed for
operation (days).
𝑇
Suppose availability is analyzed under sub-headings such as line inspection time (𝐴𝑜 ), defect
𝐾 𝑀 𝐷
detection time (𝐴𝑜 ), required material supply time (𝐴𝑜 ), and correction time (𝐴𝑜 ). In that case, the
average availability equation (Ao) is expressed as follows.
 Maulana, A Review of RAMS Analysis Application on Railway System 73

𝑇 𝐾 𝑀 𝐷
𝐴𝑜 = 𝐴𝑜 . 𝐴𝑜 . 𝐴𝑜 . 𝐴𝑜 (62)
The study conducted by [19], indicates that the availability (A) of railway systems is subject to
the influence of the mean time between failures (MTBF) and the duration required for repairing a track
post a failure occurrence. To assess the reliability, availability, maintainability, and safety (RAMS)
components, it is imperative to measure the mean time to repair (MTTR). The MTTR is contingent
upon various external maintenance resources, such as competent personnel, advanced tools and
technologies, and specialized expertise. A higher number of maintenance personnel and tools, along
with their skill levels, translates into a higher availability. Additionally, the availability is also
impacted by traffic volume, which results in a decrease as the traffic increases. The proposed model
employs the normalization of MTTR through the utilization of a function as follows:
𝐸𝑅
−τ
32𝑛
𝑎3−𝑏3.𝑒
𝑀𝑇𝑇𝑅 = 1 − −
𝐸𝑅
τ
(63)
3𝑑
𝑐3−𝑑3.𝑒

where ER = the coefficient which considers the contribution of external resources, 𝑎3 = 𝑏3 = 𝑐3,
while τ3𝑛, τ3𝑑, 𝑐3, 𝑑 , are calibrate coefficients. Under certain assumptions, the availability can be
3
expressed as follows:
𝑀𝐷𝐵𝐹
𝐴 = 𝑀𝐷𝐵𝐹+𝑀𝑇𝑇𝑅 (64)
Similarly to reference [26], reference [2] calculates availability using the time the line is open
and closed for operation. However, when there are multiple reasons (n) for intervention, such as defect
detection, material supply, and correction, the overall operational availability (A) is represented by the
following equation:
𝑛
𝐴 = ∏ 𝐴𝑖 (65)
𝑖

If the system is defined as a series, then the overall availability is:

𝑛
𝐴𝑖 = ∏ 𝑎𝑖𝑗 (66)
𝑖

For parallel system overall availability is:

𝑛
𝐴𝑖 = 1 − ∏ (1 − 𝑎𝑖𝑗) (67)
𝑗
𝑡ℎ 𝑡ℎ 𝑡ℎ
where 𝐴𝑖: 𝑖 section Overall availability, 𝑎𝑖𝑗: 𝑗 component availability at 𝑖 section, 𝑛: component
number.
Based on the study by [27], the availability is determined by calculating the mean time to
failure (MTTF) based on reliability and the mean time to repair (MTTR) based on maintainability.
𝑀𝑇𝑇𝐹
𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦 = 𝑀𝑇𝑇𝐹+𝑀𝑇𝑇𝑅 (68)
This measurement is suitable for systems that exhibit a binary state (operational or failed) and where
the comprehensive understanding of component or asset failure impact on the system is considered.

3.3.3. Maintainability of Railway Infrastructure

Maintainability is a term that denotes the ability of an object to be conserved or rejuvenated to
a state where it can function as necessitated under particular conditions of operation and upkeep. [30].
Furthermore, it will be discussed related to the study of the application of maintainability in railway
infrastructure that has been carried out by previous researchers.
74 Jurnal Teknologi dan Manajemen, Vol 4, No 2, Juli 2023: 57-77

Reference [26] explained that maintenance activities carried out at a certain time are
calculated to keep the line at the expected quality. The Poisson distribution is used in obtaining the
probability of maintenance activities, which is as follows.
−λ𝑇𝑖
(λ𝑇𝑖)𝑥.𝑒
𝑃(𝑥) = 𝑥!
(69)
where, 𝑃(𝑥): the probability of the desired amount of maintenance in a given period, λ: the amount of
maintenance performed in the selected period, 𝑇𝑖: the desired period (365 days), dan 𝑥: the probability
of occurrence of the required amount of maintenance.
Reference [19] explained that maintainability of a system is affected by two factors: the
amount of traffic it receives (which reduces maintainability as traffic increases) and the time required
for repairs, which is linked to the resources available for the task. Maintainability is defined by the
following equation:
𝑀 = 𝑀𝑇𝐵𝑀. (1 − 𝑀𝑇𝑇𝑅) (70)
Mean time between maintenance (MTBM) is assumed to be a traffic function as follows:
𝑀𝐺𝑇
− τ
4𝑛
𝑎4−𝑏4.𝑒
𝑀𝑇𝐵𝑀 = 1 − −
𝑀𝐺𝑇
τ
(71)
4𝑑
𝑐4−𝑑4.𝑒

where 𝑎4 = 𝑏4 = 𝑐4, while τ4𝑛, τ4𝑑, 𝑐4, 𝑑 , are coefficients to calibrate.

4
From reference [2], it can be inferred that probability distribution functions hold significant
value in the realm of maintenance analysis. The Poisson distribution is a frequently utilized discrete
probability distribution, as demonstrated in reference [26], however, in this study it is explained if the
system is defined as a series, then the overall maintainability is:
𝑛
𝑀𝑖 = ∏ 𝑚𝑖𝑗 (72)
𝑖

For parallel system overall maintainability is:

𝑛
𝑀𝑖 = 1 − ∏ (1 − 𝑚𝑖𝑗) (73)
𝑗
𝑡ℎ 𝑡ℎ 𝑡ℎ
where 𝑀𝑖: 𝑖 section overall maintainability, 𝑚𝑖𝑗: 𝑗 subsection maintainability in 𝑖 section, 𝑛:
subsections number.
According to a study conducted by [27], the concept of maintenance can be precisely
described mathematically as the probability that a component or system will be fully functional within
a specified time interval. This approach is analogous to the approach adopted by reference [18] when
analyzing the railway network.

3.3.4. Safety of Railway Infrastructure

Safety is the absence of unacceptably high risk [30]. Safety depends on track structure and
geometry, train traffic, and speed [19]. It also depends on maintainability. Train accidents are
dominated by track defects. Defects can be categorizes into structural failures (rails, fasteners,
sleepers, and ballast) and track geometry failures (alignment and gauge). This equation can be used to
represent safety:
𝑉
−τ
5𝑛
4 𝑎5−𝑏5.𝑒
𝑆= 1− −τ
𝑉 . 𝑀. 𝑇𝑆. 𝑆𝐸 (74)
5𝑑
𝑐4−𝑑4.𝑒

where, M is maintainability, TS is the track safety correlated with the speed effect (SE) which is
defined as:
 Maulana, A Review of RAMS Analysis Application on Railway System 75

𝐾
𝑇𝑆 = 100
(75)
where K is the coefficient, which, depending on the track's circumstances, can range from 0 to 100.
The normalized safety of the track (S) is affected by high K (ideal circumstances, TS close to 1) and
low K (unsatisfactory conditions, TS close to 0). The SE looks like this:
𝑉
−τ
5𝑛
𝑎5−𝑏5.𝑒
𝑆𝐸 = 1 − −τ
𝑉 (76)
5𝑑
𝑐4−𝑑4.𝑒

where, V is line speed in km/h, 𝑎5 = 𝑏5 = 𝑐5, while τ5𝑛, τ5𝑑, 𝑐5, 𝑑 , are coefficients to calibrate.
5
The management of a railway infrastructure manager (IM) places utmost importance on safety,
as per the findings of reference [27]. To evaluate safety performance, several performance indicators
can be utilized, including persons seriously injured and killed, significant accidents, suicides, and
attempted suicides, as well as workforce accidents. The hazardous events, identified by the RSSB are
considered within the safety parameter. The proposal suggests evaluating the outcome of each
hazardous event in relation to the fatalities and weighted injuries (FWI) index. Under the FWI index, a
fatality is assigned a score of 1, a major injury is assigned a score of 0.1, and a minor injury is
assigned a score of 0.05. Safety is evaluated as the projected number of FWI per year per kilometer.

3.3.5. Summary of RAMS Application in Railway Infrastructure

The review in this section obtained the following results which can be seen in the Table 8.
Table 8. Assessment Method for RAMS Used for Railway Infrastructure in Included Studies
No Study Reliability Availability Maintainability Safety

Hidirov and Guler Probabilistic 𝑇𝑖 Probabilistic criteria

1 𝐴𝑖 =
[26] criteria 𝑇𝑖+𝑡𝑖 (Poisson)
Maintainability,
Praticò and Giunta
2 MTBF. MDBF MDBF, MTTR MTTR, MTBM Track Safety (TS),
[19]
Speed Effect (SE)
Hidirov and Guler Probabilistic 𝑇𝑖 Probabilistic criteria
3 𝐴𝑖 =
[2] criteria 𝑇𝑖+𝑡𝑖 (Poisson)
Fatalities and
SAFs, mean time Probabilistic criteria,
4 Litherland et al. [27] MTTF, MTTR Weighted Injuries
between SAFs MTTR
(FWI)

The findings indicate that probabilistic standards are commonly employed to derive
dependability due to the variability in the frequency of infrastructure failures. Thus, this approach is
deemed appropriate for determining the reliability of a system. It is worth noting that infrastructure is
tantamount to maintenance undertakings aimed at preserving dependability, and hence, maintenance
standards serve as the principal benchmark for determining the availability. Consequently,
probabilistic standards are extensively utilized to achieve maintainability since maintenance activities
are intricately linked to the possibility of a system breakdown. With respect to the safety facet, the
methodology is tailored to suit the object of inquiry, and hence, the approach utilized differs across
studies.

4. Results and Discussions

From the review of several papers on the application of RAMS to railway systems, it is found
that to calculate reliability if the failure rate is constant, equations (1) and (2) can be used for all
systems, in this case, the railway network, rolling stock, or infrastructure. However, if the failure rate
is not constant, then a statistical model is more suitable to be applied which will be fitted to get which
model is most suitable for a system or component.
Furthermore, in calculating availability, the maintenance parameter is most widely used in
calculations, especially in railway rolling stock and infrastructure, this is because maintenance
76 Jurnal Teknologi dan Manajemen, Vol 4, No 2, Juli 2023: 57-77

activities make the system or component inoperable so that it is categorized as downtime. As for the
railway network, availability is calculated based on the overall downtime of the system, as well as the
efficiency and capacity parameters of the network. In this railway network, it is not so related to the
physical component, but rather to the arrangement or management of train travel.
Maintenance activities are the main factor in getting the maintainability of the railway system,
both preventive and corrective. In the railway network and rolling stock, MTTR is the most widely
used parameter because in these systems planned maintenance activities are often applied. Whereas in
railway infrastructure, probabilistic criteria are widely used, due to the uncertainty of maintenance
activities due to unpredictable material or system failures, so more corrective maintenance is carried
out, in addition to preventive maintenance on the system.
For the safety aspect, the application is adjusted to the system under review. There are several
different methods for obtaining safety values on the railway system, but it is all inseparable from the
reliability, availability, and maintainability of the system.

5. Conclusions
A review of previous and current research on the application of reliability, availability,
maintainability, and safety (RAMS) to railway systems is conducted and discussed in this contribution.
The methods used in calculating and obtaining RAMS values are presented and discussed in detail.
In railway networks, an assortment of maintenance parameters is employed to derive
reliability, availability, and maintainability values. Meanwhile, time and accident parameters are
utilized to obtain reliability, maintainability, and safety. Additionally, availability relies on network
capacity and efficiency parameters. It is widely known that the Weibull distribution is a popular
method used for calculating reliability within the railway industry. This is attributed to the
non-constant failure rate of railway systems, which necessitates the application of appropriate
approaches to derive reliable values. Moreover, it is worth noting that almost all maintenance criteria
are employed to achieve availability and maintainability as a direct result of reliability. The assessment
method with probabilistic criteria is the most widely used method for calculating reliability in railway
systems, especially infrastructure. This is due to the uncertainty or inconstancy of the failure rate of
infrastructure assets such as components and railroad geometry. Asset reliability is closely related to
the failure rate so it will affect maintenance and availability activities. Therefore, the method to obtain
the maintainability value of an asset is related to its failure rate, and the availability value is obtained
from the total operating time and maintenance activities carried out. This safety aspect method is
related to the characteristics of the existing railway in a study.
Currently, there is no research related to the application of RAMS in Indonesian railways even
though RAMS regulations and standards already exist and apply. Assessment of reliability,
availability, ease of maintenance, and safety integrity needs to be carried out on the Indonesian railway
system, especially related to infrastructure because currently, infrastructure is the largest contributor to
train accidents in Indonesia. Some of the methods used in previous studies may be used and applied in
Indonesian railways, where the methods to be used can be validated in advance with the characteristics
of the railways and compared with the characteristics of railways in Indonesia. The application of
RAMS is expected to improve reliability and safety and optimize the life cycle of Indonesian railway
assets.

Acknowledgments
Financial support in the form of college scholarships is provided by PT Kereta Api Indonesia
(Persero). Very useful consultative discussions with lecturers from the Faculty of Mechanical and
Aerospace Engineering, Institut Teknologi Bandung in preparing this journal review.

References
[1] Undang-Undang No. 23, “Tentang Perkeretaapian,” Lembaran Negara RI Tahun, no. 23. 2007.
[2] S. Hidirov and H. Guler, “Reliability, availability and maintainability analyses for railway
 Maulana, A Review of RAMS Analysis Application on Railway System 77

infrastructure management,” Struct. Infrastruct. Eng., vol. 15, no. 9, pp. 1221–1233, 2019, doi:
10.1080/15732479.2019.1615964.
[3] S. I. SNI IEC 62278, “" SNI IEC 62278:2002 (Ditetapkan oleh BSN tahun 2019) Standar
Nasional Indonesia Aplikasi perkeretaapian-Spesifikasi dan demonstrasi RAMS (reliability,
availability, maintainability and safety),” vol. 2002, 2002, [Online]. Available: www.bsn.go.id
[4] N. Rhayma, P. Bressolette, P. Breul, M. Fogli, and G. Saussine, “Reliability analysis of
maintenance operations for railway tracks,” Reliab. Eng. Syst. Saf., vol. 114, no. 1, pp. 12–25,
2013, doi: 10.1016/j.ress.2012.12.007.
[5] L. F. Caetano and P. F. Teixeira, “Availability approach to optimizing railway track renewal
operations,” J. Transp. Eng., vol. 139, no. 9, pp. 941–948, 2013, doi:
10.1061/(ASCE)TE.1943-5436.0000575.
[6] B. Gluzberg, V. Korolev, I. Shishkina, M. Berezovsky, P. Tregubchak, and N. Zverkova,
“Reliability indicators of railway joints and crossings,” IOP Conf. Ser. Mater. Sci. Eng., vol.
918, no. 1, 2020, doi: 10.1088/1757-899X/918/1/012147.
[7] J. G. Hwang and H. J. Jo, “RAMS management and assessment of railway signaling system
through RAM and safety activities,” 2008 Int. Conf. Control. Autom. Syst. ICCAS 2008, pp.
892–895, 2008, doi: 10.1109/ICCAS.2008.4694619.
[8] A. P. Patra and U. Kumar, “Availability analysis of railway track circuits,” Proc. Inst. Mech.
Eng. Part F J. Rail Rapid Transit, vol. 224, no. 3, pp. 169–177, 2010, doi:
10.1243/09544097JRRT296.
[9] S. D. Bemment, R. M. Goodall, R. Dixon, and C. P. Ward, “Improving the reliability and
availability of railway track switching by analysing historical failure data and introducing
functionally redundant subsystems,” Proc. Inst. Mech. Eng. Part F J. Rail Rapid Transit, vol.
232, no. 5, pp. 1407–1424, 2018, doi: 10.1177/0954409717727879.
[10] A. Morant, A. Gustafson, and P. Söderholm, “Safety and Availability Evaluation of Railway
Signalling Systems,” Lect. Notes Mech. Eng., pp. 303–316, 2016, doi:
10.1007/978-3-319-23597-4_22.
[11] M. Leite, M. A. Costa, T. Alves, V. Infante, and A. R. Andrade, “Reliability and availability
assessment of railway locomotive bogies under correlated failures,” Eng. Fail. Anal., vol. 135,
May 2022, doi: 10.1016/j.engfailanal.2022.106104.
[12] Y. H. Li, Y. D. Wang, and W. Z. Zhao, “Bogie failure mode analysis for railway freight car
based on FMECA,” Proc. 2009 8th Int. Conf. Reliab. Maintainab. Safety, ICRMS 2009, pp.
5–8, 2009, doi: 10.1109/ICRMS.2009.5270253.
[13] M. Szkoda, “Analysis of Reliability, Availability and Maintainability (RAM) of SM48 Diesel
Locomotive,” 22nd Int. Symp. EURO-ZEL, no. 37, pp. 1–10, 2014.
[14] M. Szkoda and G. Kaczor, “Reliability and availability assessment of diesel locomotive using
fault tree analysis,” Arch. Transp., vol. 40, no. 4, pp. 65–75, 2016, doi:
10.5604/08669546.1225470.
[15] Z. Zhang, L. Jia, and Y. Qin, “RAMS analysis of railway network: model development and a
case study in China,” Smart Resilient Transp., vol. 3, no. 1, pp. 2–11, 2021, doi:
10.1108/srt-10-2020-0013.
[16] J. Armstrong and J. Preston, “Balancing railway network availability and engineering access,”
Proc. Inst. Civ. Eng. Transp., vol. 173, no. 4, pp. 209–217, 2020, doi: 10.1680/jtran.19.00045.
[17] F. Nazarizadeh, A. Alemtabriz, M. Zandieh, and A. Raad, “An analytical model for reliability
assessment of the rail system considering dependent failures (case study of Iranian railway),”
Reliab. Eng. Syst. Saf., vol. 227, no. January, p. 108725, 2022, doi:
10.1016/j.ress.2022.108725.
[18] J. Litherland, G. Calvert, J. Andrews, S. Modhara, and A. Kirwan, “An Alternative Approach
to Railway RAMS: Development of an Extended RAMS Framework,” Infrastruct. Asset
Manag., pp. 1–10, Feb. 2021, doi: 10.1680/jinam.20.00002.
[19] F. G. Praticò and M. Giunta, “Proposal of a Key Performance Indicator for Railway Track
Based on LCC and RAMS Analyses,” J. Constr. Eng. Manag., vol. 144, no. 2, pp. 1–10, 2018,
doi: 10.1061/(asce)co.1943-7862.0001422.
78 Jurnal Teknologi dan Manajemen, Vol 4, No 2, Juli 2023: 57-77

[20] A. Morant, M. Westerberg, and P. O. Larsson-Kråik, “Knowledge management in a railway

network: The case of signalling systems,” Civil-Comp Proc., vol. 104, 2014, doi:
10.4203/ccp.104.274.
[21] K. R. Li Wang, Min An, Yinggui Zhang, “RAILWAY NETWORK RELIABILITY ANALYSIS
BASED ON KEY STATION IDENTIFICATION USING COMPLEX NETWORK THEORY:
A REAL-WORLD CASE STUDY OF HIGH-SPEED RAIL NETWORK,” 2017, [Online].
Available: https://usir.salford.ac.uk/44058/
[22] B. Egamberdiev, K. Lee, J. Lee, and S. Burnashev, “A Study on Life Cycle Cost on Railway
Locomotive Systems,” Int. J. Railw., vol. 9, no. 1, pp. 10–14, 2016, doi:
10.7782/ijr.2016.9.1.010.
[23] N. N. Baharin, S. H. Sanusi, N. A. Abdul, and A. A. Jumiran, “Intercity Rail Maintainability
Analysis : A Case Study on Rolling Stock Maintenance Compliance,” vol. 1, pp. 50–58, 2022.
[24] M. . Bose, D., Ghosh.G., “Measurement and Evaluation of Reliability, Availability and
Maintainability of a Diesel Locomotive Engine,” Int. J. Sci. Res. Publ., vol. 3, no. 9, pp.
714–732, 2013.
[25] M. Szkoda, “Assessment of Reliability , Availability and Maintainability of Rail Gauge Change
Systems Ocena nieuszkadzalności , gotowości i podatności utrzymaniowej kolejowych
systemów przestawczych *,” Maint. Reliab., vol. 16, no. 3, pp. 422–432, 2014.
[26] S. Hidirov, “Demiryolu A ltyapısında Güvenirlik , Bulunabilirlik , Sürdürülebilirlik ve Emniyet
( RAMS ) Özet ancak ilk yapım maliyetleri çok yüksektir . Bu sebeple ulusal demiryolu
organizasyonları yapım,” vol. 2017, no. September, 2017.
[27] J. Litherland, G. Calvert, J. Andrews, S. Modhara, and A. Kirwan, “An alternative approach to
railway asset management value analysis: application to a {UK} railway corridor,” Infrastruct.
Asset Manag., vol. 8, no. 2, pp. 85–97, Jun. 2021, doi: 10.1680/jinam.21.00003.
[28] A. P. Patra, U. Kumar, and P. O. L. Kråik, “Availability target of the railway infrastructure: An
analysis,” Proc. - Annu. Reliab. Maintainab. Symp., 2010, doi: 10.1109/RAMS.2010.5448035.
[29] S. Qiu, M. Sallak, W. Schön, and Z. Cherfi-Boulanger, “Availability assessment of railway
signalling systems with uncertainty analysis using Statecharts,” Simul. Model. Pract. Theory,
vol. 47, pp. 1–18, 2014, doi: 10.1016/j.simpat.2014.04.004.
[30] EN 50126‑1:2017, “EN 50126‑1:2017 Railway Applications - The Specification and
Demonstration of Reliability, Availability, Maintainability and Safety (RAMS),” BSI Stand.
Ltd. 2017, pp. 86–88, 2017.

How to cite this article:

Maulana A A, Suweca I W. A Review of RAMS Analysis Application on Railway System. Jurnal
Teknologi dan Manajemen. 2023 Juli; 4(2):57-77. DOI: 10.31284/j.jtm.2023.v4i2.4523