Growing demands of multiple business units in data management and data analysis required the existing data warehouse to be replaced by a new enterprise data warehouse.



Implementation of an evolutionary data warehouse with a data warehouse generator as its core, which automates recurring tasks.



Significant time and cost savings in implementation processes and compliance with standards.

About ÖBB

ÖBB-Personenverkehr AG is Austria's largest mobility service provider. With over 4,000 trains and 30,000 bus routes available, every day, more than one million passengers choose to travel on ÖBB’s rail and road transport services.



„For ÖBB Personenverkehr AG, the premise was to find a reliable strategic partner in the area of data warehousing such as Trivadis. As part of our BI modernization and consolidation program, their evolutionary approach has demonstrated significant benefits.“

Christoph Schmutz
CIO at ÖBB Personenverkehr

challenge-1080x1080 1


ÖBB Personenverkehr AG, Austria's largest mobility service provider, has terabytes of data to process: more than one million passengers are transported every day with over 4,000 trains and 30,000 bus routes.

To make traveling more contemporary and even more enjoyable for their customers, the ticketing system has been completely redesigned for all sales channels. The results are innovative digital services such as online ticketing and mobile apps.

In the background of this system, the entire business logic and thus the data structure of the interlocking systems has been revised. In order to measure or predict the utilization of the trains, for example, the ticket purchase for a particular route, as previously recorded, is no longer relevant, but the ticket purchase for a particular train. Numerous new KPIs and quality standards were introduced, such as the degree of completion of disturbances and the proportion of trains with suitable WIFI or children's cinema.

Obviously, this conversion would result in terabytes of data, which must be prepared, structured and made available to the departments for evaluation and control purposes. But neither the existing data warehouse nor the modernized IT system architecture could meet the new requirements any longer.

This circumstance led ÖBB to the BI and data warehouse expert Trivadis in Austria with the goal of building an entirely new Enterprise Data Warehouse (EDW), which can meet the new requirements and, above all, fulfill one essential requirement: the EDW solution must deliver reporting basics within a very short time and pursue an evolutionary approach as agile as possible. This led to the extent that the "E" in EDW now stands for EVOLUTIONARY.


There are three different approaches to developing data warehouses:

First, following the "business-driven" approach, the focus is on the information requirements of the end users, i.e. technical requirements of the DW are determined in the context of requirements engineering and KPIs. Based on this, first a business data model is developed, then a technical data model and eventually all the structures of the DW up to the opening up of the required data sources.

Secondly, "data-driven" is the exact opposite approach. After analysis of the available data sources, the DW is implemented according to the technically possible links. Through the various layers of a DW, the information is integrated and provided to the end user. There are two major drawbacks, though: On one hand, there is the enormous breadth of information, because all available data sets are being extracted. On the other, the risk exists that the end users might need the information provided in a different form.

The third and less well-known approach is the "method-driven" model, combining the first two approaches. Following this approach, established and proven DW architecture blueprints are used and then customized to the specific needs of the business units. This was the basis for the evolutionary data warehouse (EDW).



Growing demands of the multiple business units throughout ÖBB Personenverkehr's organization posed a major challenge to the existing data warehouse solution. A fast integration of several new data sources (e.g. loyalty card system, web shop and ticketing system, revenue management, etc.) was required.

Trivadis chose the evolutionary approach as a project approach, which primarily involves an iterative implementation of small project contents with a high degree of automation in short release cycles.

First, a new BI Competence Center was established. The requirements for the new data warehouse solution were collected and defined in close cooperation with the business units. The implementation and integration of several new data sources took place in short iterations and pursued the agile development methodology. Benefiting from advanced data warehouse automation, the development effort and time-to-deliver were reduced to a minimum.

"The agile project approach will continue to prevail in data warehouse development. Additionally, automation, especially with regard to data architecture and data transformations, provides decisive advantages. In addition to the strong business-oriented methodology, Trivadis brings the technical standards along with tested and proven best practices into development projects" explains Günther Krobath, Solution Unit Manager for Business Intelligence at Trivadis in Vienna.



Traditional data warehouse solutions can no longer keep up with the shorter intervals of new requirements. Until they are implemented, the need for information has mostly changed or even been eliminated and new aspects have been added. With standardization and automation, however, DW processes are gaining in effectiveness as significant time savings and cost advantages can be realized. The core of an evolutionary data warehouse is a data warehouse generator, which automates recurring tasks.

This results in immense time savings in implementation, e.g. the process can be completed within hours or within few days - from the definition of the requirements in the business units to their implementation. The use of a generator brings another positive side effect: it ensures compliance with standards and far less testing effort. The critical factor for success is, how flexibly the generator can be customized to the desired DW architecture and if the entire business logic for processing the data can be managed in the tool. High level of customizability of the generator is the premise to ensure the advantages of automation across the entire lifecycle of a data warehouse.


  • SAP Business Objects
  • Oracle
  • biGENIUS®


ÖBB Personenverkehr AG, Transport & Logistics


In a close co-operation of .BB Personenverkehr with Trivadis, the basic methodology was introduced at the beginning of the project and the departments were involved in the iterative process very early on.

"Evolutionary data warehousing enabled the desired short time-to-market strategy in close business interaction. Demands could be transformed and presented in results in a very short time. Here, Trivadis gave us the best possible support and contributed their many years of expertise to the processes", reports Roman Novak, Chief Information Architect .BB Personenverkehr.