Unseen Connections — Indian Railway Junctions
This is a first attempt to create my own story and I have chosen to try on my favorite hobby of travelling through Indian Railways. Being a avid traveler and a User Experience professional I have been observing some gaps which railway commuters are encountering while they travel. Out of many in this post I am trying to highlight unseen connections of daily trains at Vijayawada (Bezawada — second busiest railway junction in India) railway junction in Andhra Pradesh which is always a puzzle for a common commuter, and this puzzle carry forward to all the railway junctions in India.
At Industrial Design Center of Indian Institute of Technology Bombay (
), an attempt has been made to create some sense out of the available Indian railways data of daily trains at Vijayawada junction through data visualization of connected trains based on directions of destination of the train as a medium.Note — This post consists of steps involved in learning and exploring the data visualization and tools. But provides some insights for further refinements of railway connected trains.
Learnings from Knowledge Bank
Railways has been fascinating for many since decades and it is also a challenging in organizing and drawing conclusions from the huge railway data. Indian railways is the fourth largest network and the passengers comprising 119,630 kilometers (74,330 mi) of total track[4] and 92,081 km (57,216 mi) of running track over a route of 66,687 km (41,437 mi) with 7,216 stations at the end of 2015–16. There are huge knowledge bank available on railway networking visualisation, few of them are mentioned in the new few sections wherever relavant.
Mayer’s Chart is one of the brilliant visualization introduced to me by @Prof. Venkatesh Rajamanickam in the data visualisation module in
. This has inspired me to explore further on Indian Railways and specific to Vijayawada railway junction. It went on to some research on types of railway network chats and blogs which has provided some basics on how to start visualising the data. Few of them Mayer’s chart, London Tube network, Indian Railway live train mapping, Socialcop, Mapbox, Tableau, Google maps, open rail maps, more in references section.Few blogs
Ten Examples of the Subway Map Metaphor
The Best & Worst Subway Map Designs From Around the World
An interactive exploration of Boston’s subway system
Data Source
Initial data source of Indian Railways was taken from data.gov of the year 2015.
But it is found have many insufficient daily trains schedule data for Vijayawada Railway station, so eventually the other sources to looked at are Trains at glance by Indian Railways, IRCTC, Cleartrip, etrains, indiarailinfo for data mining. The data that was captured was
Train Name
Train Number
Daily or Weekly
Arrival Time
Departure Time
Stoppage Time
Most of the trains data was up to date as they deal with daily reservations of millions of commuters. So this data is take forward for data analysis.
Data Analysis
Only daily trains at Vijayawada junction was taken as a sample data for analyzing the data. The mined data was compared manually and analysed with the tools like MS Excel and Tableau to cross check on the quality of data. To enrich the data for creating of good patterns few more data was derived from the existing data, which are
Source Station
Destination Station
Coming From (North, East, West, South)
Going Towards (North, East, West, South)
Trains that are Starting at Vijayawada
Trains that are Ending at Vijayawada
Day or Night train
Note — For the convenience of analysis only four geographical poles are considered.
Deriving and analysing data along with these fields made little easier to see some sensible patterns or models in arrivals and departures at this station.
The patterns are like most of the traffic is from East, South and West for these daily trains. All the long journey trains are between North and South which are weekly trains. This evidence given some thoughts on data visualization ideas. Also all the intercities are clearly marked which connects East, West and South majorly at this junction.
Making Sense out of Data
You can download this data for further analysis of other data visualisation.
This data was considered for further creating of data visualisation ideas using various tools.
Drawing insights from the below charts to understand from which stations the inflow is more and similarly the outflow for Vijayawada station.
It is evident that traffic between Visakhapatnam (East), Secunderabad (West) and Chennai (South). And there 6 inter-city express trains which ply from and to Vijayawada connecting East, West and South bound cities.
Now to analyse the rush hours and train available we need to drill down much more into analysis. For this Tableau is used to create more complex connections.
Tool Selection
Being a non coder, I have chosen MS Excel and Tableau to explore the concepts and for different visualisations. Tableau needs a simple Excel or a .CSV file to start with. After importing, it will analyse the data fields and items and categories them as Dimensions and Measures. Dimensions are basicially with have parameter to create combinations for visualisation and Measures have values which are mostly numerics and can be added to sum up, mulitiply, etc. Its any easy tool for the basic charts creation, to advance in customising the graphic you may need some programming skills as well, which is out of scope for me.
Another tool which I have used for custom visualisation is Adobe Illustrator.
Keep it simple — Data Vs Visualisation
After data got imported here it comes Data vs. Chart. So going back to basics of research knowledge, but definitely Tableau has helped me by its intelligence to reduce the cognitive load on me and decision making. According to the type of data the tool suggested few charts or combination of charts to visualize the right things in easiest way.
Few References
The Extreme Presentation(tm) Method or Home page
Which chart or graph is right for you?
Dissecting-how-to-choose-the-right by Qlik
How to Choose the Right Chart or Graph for Your Data
Analysis and Interpretation
Grouping of information in a meaningful way (refer Gestalts laws) is very important here and one needs to understand the data very well to derive some useful and informative visualization. Initially it was a struggle to derive any visualization to see if it makes any sense, it also a phase where Tableau was explored to experiment. As mentioned earlier in the intro of Tableau it is important to understand the dimensions and measure, since in this data the measures were almost nil, so it was becoming difficult to create combinations.
The following visualizations are created on the following basis
- Towards Journey
- Arrival time
- Departure time
- Stoppage Time
- Destinations
Initially the attempt was made to find number of trains that were bound towards North, East, West, South, Start and End, and the same data is
with destinations.
From the above two visualisations it is clear that there 6 trains which are ending at Vijayawada and more number of trains are plying towards East and West.
When applied some more dimensions like train number and name, to visualize in cluster the below pattern evolved.
This shows some outliers in a hexagonal shape and if we draw the boundaries it shows the below.
This bubble chart was sorted to view the cluster of trains bounded towards each direction.
And with Day and Night measures
From the above visualization it is clear that Vijayawada station handles more number of East bound trains and they also handle most trains in the night time. Interestingly most of these outliers are the trains which ends at Vijayawada.
Now to view by destination the data is visualised through Tree map
Another version of tree map according to the direction of travel.
As the aim of this exercise is to find connected trains, time is introduced on the top of the above tree map and sorted by time.
This gives me an overall picture of trains in which direction the trains are bounded and time for connections.
Still the above Tree map may not be a quick way to interpret the connections the exploration went on.
Here there are few combinations of measures are tried to see if there is any relationship between Train departures, number, and bounds of Arrival and Departure.
In the above connections it is more clear on which train is arriving from which direction and what are the connected trains towards which direction. In this format with the help of filters specific time based trains can be found.
The above visualisation illustrates stoppage time against departures. The null row illustrates the trains either starts or ends in Vijayawada junction.
To more simply here is the chart provides all sequential connected trains.
With the combinations of the filters one can find which trains that they can board for a particular time.
The purpose of the above analysis is for deriving a good info graphics for Vijayawada station. Which will be posted in the next article. Below image provides a taste of it.
Here are the published interactive Tableau online links for the data visualizations
I am sure there are much always better ways to design, so open for feedback and suggestions.
More References
Data Visualization Tools
Book on Cartographies
Project Mapping
Tableau White paper
Google Search Keywords
Rail network, London national rail network, tube map, Singapore rail, train maps, metro rail maps, Dubai metro map, chose the correct chart, data vs chart.
Thank You!
