Presentation on Unleashing SAP HANA capabilities in Indian Railways

1. Prateek Tyagi2. Tanuj DoshiNTT Data GDS

Unleashing the Capabilities of SAP HANA in Indian Railways

What is HANA?

Why HANA?

HANA Capabilities?

SAP HANAHigh Performance Analytic ApplianceIt’s a platform, but not an application, but its an

applianceHybrid DatabaseColumnar StorageIn-Memory Data StorageSoftware InnovationsHardware InnovationsManaging big data has become lot more easierReal time analysis at unprecedented speed on large

volumes of dataSAP HANA editions

SAP HANA ARCHITECTURE

Key PointsData Provisioning ETL Tools with SAP HANA

SAP SLTSAP BODSDXCManual loading using flat files using HANA

StudioSybase Replication Server

Replication TechnologiesTrigger Based ReplicationETL Based ReplicationLog Based Replication

SAP HANA DB ServerName ServerIndex ServerStatistics

ServerPre Processor

Why This TopicSome Facts about Indian Railways:

Rail Network: 115,000 kmTrack Route: 65,436 kmNo. of Stations: 7,172Passengers Carried: 23 million dailyNo. of Passenger Trains: 12,617

Looking at such heavy and facts, its obvious that huge data is being generated every second of the day. This data has to be used systematically in order to gain valuable insights

After looking at the capabilities of SAP HANA, these millions of entries could be analyzed effectively and efficiently

Present Scenario• Overall Indian Railways sell 21 million tickets on an average every day. 250 million tickets are booked via IRCTC per annum. So, when booking window opens at 8.00 AM there is a mad rush for tickets both online as well as in Railway reservation counters both vying for database resources. Railway counters book from a separate system which is directly connected to the back-end infrastructure and hence would not face the issue faced by online users.

• The problem of running reports and ad-hoc queries is faced by data administrators at the back-end who often fail to deliver appropriate decisions at required time because of the queries being run on transaction system.

• Data Warehouse Application is divided into two parts front-end and back-end. Front-end is developed in ASP. At the backend there is Sybase IQ (as database), which is used to store information and deliver information to the front-end.

CHALLENGES The challenge confronting Indian Railways is to manage vast operations, get right information

at right time to make key decisions:

As the online PRS generates increasing volume of data, the need to analyze data is becoming more and more by higher management for

Managing and utilizing the existing resources Creating new resources for the convenience of passengers. Railway managers (even the Railway Programmers) cannot generate ad hoc

queries. Prevent analytical users from running queries ex: how many passengers are

booked for a given journey day.

As these slow down the performance of mission-critical PRS databases. This is because the application database is segmented across multiple servers throughout the country making it difficult for railway users to access the data on the fly.

OUR PROPOSALSAP HANA as a differentiatorThe data from PRS can be imported into SAP HANAThen we propose to create views pertaining to the

KPI’s that we want to analyze, for ex: how many passengers are booked for a given journey day

Once these views are created, we can consume/analyze these views over SAP Lumira and create meaningful visualization that will pass on the information required by the data administrators

The data from the SAP HANA can be used to do predictive analysis to estimate the passenger traffic and this will help the key decision makers to be ready with the essential requirements to fulfill the demand.

Business ScenarioReducing the Revenue Lost:

Quite a number of seats will be in waiting list even after the train has departed.

This results in returning the money booked to the passengers who could not travel in that particular train.

So why not carefully analyze the no. of waitlist passengers and then accordingly attach extra coaches

In this way, the passengers would be able to travel hassle free and results in revenue increase

Lets see how much revenue could be gained if the huge data is analyzed properly through a small demo by doing a role play

WHY ONLY SAP HANA•On an average, approx. 30 million passengers travel daily with Indian Railways.

•More than 10000 trains run daily

•Data volume for a year will be approximately 10 Billion

•The speed with which the information is made available can be improved significantly

•With in-memory computing capabilities and ability to execute complex logic these issues would be solved easily

•HANA hardware will take care of the scale of transactions happening; we only need to ensure that the web app is available all times

•Consider this in current scenario, on logging in, several calls are made to the database

•Calls are made to retrieve train code, station codes, fill in passenger details, search availability, etc. HANA can do all these in parallel.

Other Business ScenariosFew other Business Scenarios or KPI’s that could

be analyzed:Total no. of passengers travelling on a particular

route on any given dayTotal no. of seats that are going vacantNo. of cancellations that happened on any given dayAdjustment of station to station quota seats within a

particular routeNo. of trains running late/in time on any given day

And many more…

Thanks

Thanks for attending

Feed us back