India has one of the largest rail networks in the world. Since the establishment of railway tracks, travelling has become easy. More than one lakh trains operate around India every single day and carry millions and billions of passengers. Indian railways are cost-effective, reliable and environment friendly that’s why it is often called as the "lifeline of the nation". Every day thousands of passengers use the Indian rail network to reach their destinations.So it is essential to ensure the safety of passengers.

Actually the term ‘accident’ envelopes a wide spectrum of occurrences with or without significant impact on the system.These consequential train accidents may be due to collisions, derailments, fire in trains, road vehicles colliding with trains at level crossings, and certain specified types of ‘miscellaneous’ train mishaps.

The country has witnessed several tragic train accidents.Recently, on 2nd June, one of the deadliest train accidents took place three trains collided in Odisha Balasore , nearly 300 passenger died and nearly 1000 injured.

The Coromandel Express collided with a stationary goods train, causing many of its carriages to overturn, some of which fell onto another train, the Bengaluru-Howrah Superfast Express, which was also passing by at the same time on 2 June 2023. (Image Source: PTI)
The Coromandel Express collided with a stationary goods train, causing many of its carriages to overturn, some of which fell onto another train, the Bengaluru-Howrah Superfast Express, which was also passing by at the same time on 2 June 2023. (Image Source: PTI)


Simplified diagram of the 2023 Odisha train collision
Simplified diagram of the 2023 Odisha train collision


-> Now we will explore the total number of passengers killed in train accidents since 2000 in India, shedding light on the safety of rail travel and the measures in place to prevent such incidents.

Number of Passengers killed and injured in train accidents

(since 2000)


library(readxl)

data1 = read_excel("/Users/khushi/Documents/EDA_Analysis_of_train_accidents/Data_of_train_accidents.xlsx",sheet='Killed')

library(tidyr)
data_pivot <- pivot_longer(data1, 
                      cols = c("Number_of_Passengers_   Killed","Number_of_Passengers_  Injured"), 
                      names_to = "Passengers_Killed", 
                      values_to = "Number")
library(ggplot2)
ggplot()+ 
geom_col(data = data_pivot, 
        aes(x=Y,y=Number,fill=Passengers_Killed), 
        position = "dodge")+
  ylab("Number of Passengers")+
  xlab("Year")+
  labs(title = "  Number of Passengers killed and injured in train accidents")

Primary Reasons behind Railway Accidents

Train accidents can be categorized into four primary types: Collisions, Derailments, and Fires. Each of these types has its own set of causes and contributing factors. Here’s an explanation of the reasons for train accidents in these categories:
1.Collisions: Train collisions are railway accidents that occur when two or more trains come into contact with each other. These collisions can have serious consequences, resulting in injuries, fatalities, and extensive damage to railway infrastructure.
2.Derailments: Train derailments are railway accidents in which a train leaves the tracks, leading to potentially disastrous consequences.
3.Fire in trains: Fires in trains are hazardous incidents that pose a significant risk to passengers, railway personnel, and railway infrastructure.
4.Level Crossings: Many railway accidents occur at level crossings where trains intersect with roads. Poorly marked or unprotected crossings, as well as driver negligence, can lead to collisions.

To know what is the main cause of accidents we took the data having these variables.

library("readxl")
data = read_excel("/Users/khushi/Documents/EDA_Analysis_of_train_accidents/Data_of_train_accidents.xlsx",sheet='Table 1')
library(ggplot2)
data = read_excel("/Users/khushi/Documents/EDA_Analysis_of_train_accidents/Data_of_train_accidents.xlsx", sheet=1)
ggplot(data) +
  geom_line(mapping = aes(x = Y, y = Collisions, color = "Collisions")) +
  geom_point(mapping = aes(x = Y, y = Collisions, color = "Collisions")) +
  geom_line(mapping = aes(x = Y, y = Derailments, color = "Derailments")) +
  geom_point(mapping = aes(x = Y, y = Derailments, color = "Derailments")) +
  geom_line(mapping = aes(x = Y, y = Fire_in_trains, color = "Fire in trains")) +
  geom_point(mapping = aes(x = Y, y = Fire_in_trains, color = "Fire in trains")) +
  geom_line(mapping = aes(x = Y, y = Level_crossing_accidents, color = "Level crossing accidents")) +
  geom_point(mapping = aes(x = Y, y = Level_crossing_accidents, color = "Level crossing accidents")) +
  scale_color_manual(values = c("Collisions" = "blue", "Derailments" = "red", "Fire in trains" = "green", "Level crossing accidents" = "yellow")) +
  labs(color = "Causes of Train Accidents")+
  ylab("number of train accidents")+
  xlab("Year")+
  labs(title = "  Causes of Train Accidents Over the Years")

we observed that the total number of accidents is higher in the case of derailments for the Years from 1970 to 2010, it suggests that derailments might be a significant cause of train accidents for this specific time period.

data5 = read_excel("/Users/khushi/Documents/EDA_Analysis_of_train_accidents/Data_of_train_accidents.xlsx", sheet='Table 12')
ggplot(data5)+
  geom_bar(aes(x="", y=Total, fill=Causes),
           stat="identity", color="white")+
  coord_polar("y", start=0)+
  ggtitle("Number of Accidents")+
  theme(plot.title= element_text(hjust = 0.5, size=20),
        axis.title = element_blank(),
        axis.text = element_blank(),
        axis.ticks = element_blank(),
        panel.grid = element_blank(),
        panel.border = element_blank())

library("VennDiagram")
## Loading required package: grid
## Loading required package: futile.logger
grid.newpage()
draw.pairwise.venn(area1 = 3076,
                   area2 = 2183,
                   cross.area = 2183,
                   fill = rainbow(2),
                   category = c("Accidents","Derailments"))

## (polygon[GRID.polygon.252], polygon[GRID.polygon.253], polygon[GRID.polygon.254], polygon[GRID.polygon.255], text[GRID.text.256], text[GRID.text.257], lines[GRID.lines.258], text[GRID.text.259], text[GRID.text.260])

Total number of accidents is 3076 for the year 2000 to 2018. Out of which 2183 are due to derailments.

Broad Causes of Train Accidents

1. Collisions:
a)Signal Failures: Malfunctioning or misinterpretation of signals can lead to collisions. Trains rely on signaling systems to maintain safe distances and directions, and any failure in the signaling system can be catastrophic.
b)Human Error: Train operators may make mistakes, such as failing to obey signals, misunderstanding instructions, or not following proper procedures. Inadequate training, fatigue, and distraction can contribute to human errors.

2. Derailments:
a)Track Defects: Poorly maintained tracks, damaged rails, and worn-out or improperly aligned switches can cause derailments. This is especially true in regions with inadequate track maintenance.
b)Equipment Failures: Mechanical failures in the train’s rolling stock, such as brake failures or wheel problems, can result in derailments.

3. Fires in Trains:
a)Electrical Faults: Short circuits, damaged wiring, or electrical system failures can lead to fires on trains, especially in older train cars with outdated wiring.
b)Sabotage and Acts of Terrorism: In some cases, railway accidents result from deliberate sabotage, including acts of terrorism. These incidents can have devastating consequences for passengers and railway infrastructure.

data4 = read_excel("/Users/khushi/Documents/EDA_Analysis_of_train_accidents/Data_of_train_accidents.xlsx", sheet='Table 13');
ggplot(data4, aes(x = "", y = No_of_accidents, fill = Broad_causes)) +
  geom_bar(stat = "identity", width = 1) +
  coord_polar(theta = "y") +
  theme_minimal() +
  labs(fill = "Broad Causes") +
  ggtitle("Number of accidents (Broad Causes of Train Accidents)")

library(gridExtra)
df= read_excel("/Users/khushi/Documents/EDA_Analysis_of_train_accidents/Data_of_train_accidents.xlsx", sheet='Table 3')

p1<-ggplot(df, aes(x = Y, y = Failure_of_Railway_Staff)) +
  geom_line(color = "blue") +
  labs(x = "Year", y = "Failure of Railway Staff")

p2<-ggplot(df, aes(x = Y, y = Failure_of_Person_other_than_Railway_Staff)) +
  geom_line(color = "red") +
  labs(x = "Year", y = "Failure of Person other than Railway Staff")

p3<-ggplot(df, aes(x = Y, y = Failure_of_Equipments_Rolling___Stock)) +
  geom_line(color = "green") +
  labs(x = "Year", y = "Failure of Equipments Rolling Stock")

p4<-ggplot(df, aes(x = Y, y = Failure_of_Equipments_Track)) +
  geom_line(color = "violet") +
  labs(x = "Year", y = "Failure of Equipments Track")

p5<-ggplot(df, aes(x = Y, y = Failure_of_Equipments_Electrical)) +
  geom_line(color = "black") +
  labs(x = "Year", y = "Failure_of_Equipments_Electrical")

p6<-ggplot(df, aes(x = Y, y = Sabotage)) +
  geom_line(color = "orange") +
  labs(x = "Year", y = "Sabotage")
grid.arrange(p1, p2, p3, p4, p5, p6, ncol=3, nrow=2)

State-wise number of Manned and Unmanned Level Crossings

There are 19,507 Manned Level Crossings and 5792 Unmanned Level Crossings in the country as on 01.04.2018.

data2 = read_excel("/Users/khushi/Documents/EDA_Analysis_of_train_accidents/Data_of_train_accidents.xlsx", sheet='Table 10');
## New names:
## • `Number_of_Manned__Level_Crossings_` ->
##   `Number_of_Manned__Level_Crossings_...3`
## • `Number_of_Unmanned_Level_Crossings` ->
##   `Number_of_Unmanned_Level_Crossings...4`
## • `Number_of_Manned__Level_Crossings_` ->
##   `Number_of_Manned__Level_Crossings_...7`
## • `Number_of_Unmanned_Level_Crossings` ->
##   `Number_of_Unmanned_Level_Crossings...8`
## • `Number_of_Manned__Level_Crossings_` ->
##   `Number_of_Manned__Level_Crossings_...9`
## • `Number_of_Unmanned_Level_Crossings` ->
##   `Number_of_Unmanned_Level_Crossings...10`
library(gridExtra)
pl1 <- ggplot()+ 
  geom_col(data = data2, 
           aes(x=State,y=MLC))+
  ylab("Number of Manned Level Crossings")+
  coord_flip()+
  xlab("States")

pl2 <- ggplot()+ 
  geom_col(data = data2, 
           aes(x=State,y=UMLC))+
  ylab("Number of Unmanned Level Crossings")+
  coord_flip()+
  xlab("States")
grid.arrange(pl1,pl2, ncol=2)

Accidents at Unmanned Level Crossings

The number of accidents/causalities, due to Unmanned Level Crossings in the country, from 2010-11 to 2017-18 and the current year 2018-19 (upto May 2018) are shown below:

library(gridExtra)
data3 = read_excel("/Users/khushi/Documents/EDA_Analysis_of_train_accidents/Data_of_train_accidents.xlsx", sheet='Table 9')
plot1<- ggplot()+ 
  geom_col(data = data3, 
           aes(x=Y,y=Number_of_accidents_at_Unmanned_Level_Crossings, col=factor(Y)))+
  ylab("Number of Accidents at Unmanned Level Crossings")+
  ggtitle("Number of Accidents at Unmanned Level Crossings")+
  xlab("Year")
data6 = read_excel("/Users/khushi/Documents/EDA_Analysis_of_train_accidents/Data_of_train_accidents.xlsx", sheet='Table 9')

plot2<-ggplot(data6, aes(x = Y, y = Number_of_accidents_at_Unmanned_Level_Crossings)) +
  geom_smooth(method = "loess", color = "red") +
  labs(x = "Year", y = " ") +
  theme_classic()
grid.arrange(plot1, plot2, ncol=2)
## `geom_smooth()` using formula = 'y ~ x'

It is the endeavor of the Ministry of Railways to eliminate all Unmanned Level Crossings, in a phased manner, by either of the following:-

  • Closure - Closing Unmanned Level Crossings having NIL/ negligible Train Vehicle Unit (TVU).
  • Merger - Merger of Unmanned Level Crossing to nearby Manned/Unmanned Level Crossing or Subway/Road Under Bridge (RUB)/Road Over Bridge (ROB) by construction of diversion road.
  • Provision of Subways/RUBs.
  • Manning- Phased manning of unmanned level crossings which cannot be eliminated by the above means.

Damage to Railway Property

Details of the damage to railway property and interruption to communications caused by train accidents since 2000

data7 = read_excel("/Users/khushi/Documents/EDA_Analysis_of_train_accidents/Data_of_train_accidents.xlsx", sheet='Table 6')
library(tidyr)
data7 <- pivot_longer(data7, 
                      cols = c("Cost_of_Damage_RollingStock","Cost_of_Damage_PermanentWay"),
                      names_to = "Damage",
                      values_to= "Rupees")
ggplot()+ 
  geom_col(data = data7, 
           aes(x=Y,y=Rupees,fill=Damage), 
           position = "dodge")+
  ylab("Damage")+
  xlab("Year")+
  labs(title = "Damage caused to Indian Railways (in Lakhs)")

Mission ‘Zero-Accident’

In the Railway Budget 2016-17, Mission Zero Accident was announced, comprising of the two sub-missions:

  1. Elimination of Unmanned Level Crossings: As on 01.04.2017, Indian Railways have 27181 Level Crossings, out of which 19480 are manned and 7701 are unmanned. Out of total 7701 unmanned level crossings, 4943 level crossings are on BG (Broad Gauge). It had been planned to eliminate unmanned level crossings (UMLCs) on Broad Gauge by 2020.

  2. TCAS (Train Collision Avoidance System): To prevent collisions and signal passing at danger by the Loco Pilot through developing an indigenous technology and also to increase throughput by increasing average sectional speed on Indian Railways (IR). In order to improve signaling, electrical/electronic interlocking system with centralized operation of points and signals are being provided to eliminate human failure and to replace old mechanical systems. Superior and safer Linke Hofmann Busch (LHB) Coaches having anti-climbing features have been progressively introduced on Indian Railways to reduce the fatalities in case of accidents.

Trend of the acidents over the year

ggplot(data) +
  geom_line(mapping = aes(x = Y, y = Total)) +
  geom_point(mapping = aes(x = Y, y = Total))+
  ylab("Accidents")+
  xlab("Year")+
  ggtitle("Total number of train accidents:
          Trend over the year")

This plot show that there is a significant decline in the number if accidents over the years. This is the result of the steps taken by the Ministry of Railways, however it remains essential to sustain these efforts and further enhance safety measure to ensure the continued decline in train accidents.

Bibliography:

  1. India. Ministry of Railways, Outcome and Performance Budget, 2016-17, pg.50.
  2. India. Ministry of Railways, Safety Performance 2014-15, URL: http://www.indianrailways.gov.in/railwayboard/uploads/directorate/safety/downlo ads/Ch1-eng-0304.PDF
  3. Information provided by Ministry of Railways vide Office Memorandum No. 2018/Safety(A&R)/15/1.
  4. Lok Sabha Starred Question No.42 dated 19.07.2017 .
  5. Lok Sabha Unstarred Question No.3013 dated 11.08.2017.
  6. Rajya Sabha Unstarred Question No. 3867 dated 09.08.2017.
  7. Rajya Sabha Unstarred Question No. 2194 dated 05.01.2018.
  8. PIB Release dated 05.01.2018.

    ………………………………………………………………………………………………………………………………………….

    In R, the sessionInfo() function is used to obtain information about the current R session.This function provides details about the version of R that is running, the operating system, the loaded packages, and various other configuration details.
sessionInfo()
## R version 4.3.1 (2023-06-16)
## Platform: aarch64-apple-darwin20 (64-bit)
## Running under: macOS Sonoma 14.2.1
## 
## Matrix products: default
## BLAS:   /Library/Frameworks/R.framework/Versions/4.3-arm64/Resources/lib/libRblas.0.dylib 
## LAPACK: /Library/Frameworks/R.framework/Versions/4.3-arm64/Resources/lib/libRlapack.dylib;  LAPACK version 3.11.0
## 
## locale:
## [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
## 
## time zone: Asia/Kolkata
## tzcode source: internal
## 
## attached base packages:
## [1] grid      stats     graphics  grDevices utils     datasets  methods  
## [8] base     
## 
## other attached packages:
## [1] gridExtra_2.3       VennDiagram_1.7.3   futile.logger_1.4.3
## [4] ggplot2_3.4.2       tidyr_1.3.0         readxl_1.4.3       
## 
## loaded via a namespace (and not attached):
##  [1] sass_0.4.7           utf8_1.2.3           generics_0.1.3      
##  [4] futile.options_1.0.1 lattice_0.21-8       digest_0.6.33       
##  [7] magrittr_2.0.3       evaluate_0.22        fastmap_1.1.1       
## [10] cellranger_1.1.0     jsonlite_1.8.7       Matrix_1.6-0        
## [13] formatR_1.14         mgcv_1.9-0           purrr_1.0.1         
## [16] fansi_1.0.4          scales_1.2.1         jquerylib_0.1.4     
## [19] cli_3.6.1            rlang_1.1.1          munsell_0.5.0       
## [22] splines_4.3.1        withr_2.5.0          cachem_1.0.8        
## [25] yaml_2.3.7           tools_4.3.1          dplyr_1.1.3         
## [28] colorspace_2.1-0     lambda.r_1.2.4       vctrs_0.6.4         
## [31] R6_2.5.1             lifecycle_1.0.3      pkgconfig_2.0.3     
## [34] pillar_1.9.0         bslib_0.5.1          gtable_0.3.3        
## [37] glue_1.6.2           xfun_0.40            tibble_3.2.1        
## [40] tidyselect_1.2.0     rstudioapi_0.15.0    knitr_1.44          
## [43] farver_2.1.1         htmltools_0.5.6.1    nlme_3.1-162        
## [46] rmarkdown_2.25       labeling_0.4.2       compiler_4.3.1