Lufthansa Systems ‘crowdfunds’ BI to mine flight data

Open source business intelligence tool enables airline to gain new insights from a decade’s worth of ticket sales

When software developers at Lufthansa Systems spotted an opportunity to use business intelligence (BI) technology to develop better information systems for airlines, they turned to their employer’s “crowdfunding” scheme to raise finance for the idea.

Analytics lead Tamas Farkas and senior architect Árpád Csordás helped their teams to win enough votes from other employees at Lufthansa Systems to secure an initial round of funding from the company – but it meant thinking like salesmen and marketers rather than technical specialists.

“The initiative came from the developers,” says Csordás. “We have more than 100 products and we saw the need to have something new – modern and centralised reporting.”

Lufthansa Systems, which supplies IT to 300 airlines, had found it difficult and time-consuming to produce complex business reports from its company data, but Farkas and Csordás knew there was a better way.

They spent several months testing software and speaking to suppliers before opting for the Jaspersoft BI package supplied by US integration and BI company Tibco.

The team built five different prototype applications, each of which took 100 hours to create, to demonstrate how BI could benefit the business.

“We are technical people, but here we had to do sales and marking,” says Csordás. “We had to convince our boss and management that there was a valid use case.”

Just the ticket

One of the projects that followed is now helping the airline Lufthansa Group, which operates as a separate company, to gain new insights from its database of ticket sales.

The airline group’s Oracle database contains details of millions of tickets stretching back 10 years, but producing complex analysis proved time-consuming. Some reports took many hours to complete and could only be run overnight.

The development team at Lufthansa Systems used agile scrum to develop the first practical demonstration for Lufthansa Group within three weeks.

“They were not prepared for that,” says Csordás. “They were used to expecting projects delivered after three months. We were showing them something new.”

More projects

Predicting the best place for servers

Lufthansa Systems is using Jaspersoft to predict which geographic regions are most frequently used by passengers.

It has helped Lufthansa Group, which hosts servers in Amazon Web Services datacentes, identify the best geographic locations to install servers to reduce the distance data has to travel before it reaches customers.

Fishing for data

A partnership between Lufthansa Systems and the German Aerospace Centre will use analytics technology to enable the scientists at the centre to analyse huge quantities of data from multiple sources, stored in a “data lake”.

The project will identify connections between diverse data sources, providing new insights for scientists.

The project, which went live in September 2016 after six months’ work, will enable Lufthansa Group to gain new insights from its ticket data more quickly.

It will also save money by allowing the group to retire its Oracle data and move it to an open source alternative, Postgres. “It’s a huge cost saving,” says Farkas.

The work required the development team to add new capabilities to Jaspersoft – something that would have been difficult or impossible to do with software based on proprietary code, rather than open source.

“The fact that it is open source was the reason we started to work with Jaspersoft”

Tamas Farkas, Lufthansa Systems

“The fact that it is open source was the reason we started to work with Jaspersoft,” says Farkas. “It is really important for us to extend the system and add new functions. For other similar tools, it is pretty hard to request a feature from the provider.”

The project has now attracted interest from other airlines in the Lufthansa Group, including Austria and Brussels airlines.

Getting to the gate in time

Lufthansa Systems went on to use its data analytics capabilities to develop a mobile app for the Lufthansa airline group that will help travellers reach the departure gate on time.

The app takes passenger data collected in MongoDB – an open source, NoSQL database – and uses it to calculate how much time each passenger has before they need to head to the gate.

“The primary goal is to support project managers to have a better insight”

Árpád Csordás, Lufthansa Systems

The app, known as inTime, is expected to bring in extra revenue by ensuring that passengers can spend time shopping or eating in restaurants without worrying about missing their flights.

InTime will take location data from mobile phones when passengers pass through security, and potentially from Wi-Fi hotspots, so it can calculate journey times.

“It will be able to give passengers advice, for example, if they have an hour of free time because their flight is delayed, they can check out this shop or this restaurant,” says Csordás.

Chasing bugs

A third project will help Lufthansa Systems produce detailed reports from the software development team’s bug-tracking system, Jira.

The company uses Jira to track the progress of thousands of software projects, but its reporting capabilities are relatively limited.

The analytics team has used Jaspersoft to develop a sophisticated reporting tool. When it goes live on 1 July, it will present project managers with a dashboard, showing where software bugs are located, how long they will take to fix, and whether each project is on track.

“It really helps at a management level,” says Csordás. “The primary goal is to support project managers to have a better insight. All of the tasks, milestones and the ‘to dos’ are kept in this reporting system.”

Read more on Business intelligence software