When data makes the difference

At the Swiss Alliance for Data-Intensive Services’ Expert Group Event «Data Driven Business Model» hosted by D ONE, Axel Obermeier and Bojan Škerlak showed how data can make an impact at two long-established companies SBB and Migros.

Data driven efficiency gains in track maintenance at SBB
SBB leads about 200 track maintenance projects per year. The average duration of the projects – including the whole process from planning to execution – takes about 4 years average. The business unit responsible for the maintenance projects is on its journey to data-driven insights and most important, optimization: regarding the prediction of booking volumes and of project cost, but also when it comes to cost attribution for interrelated drivers, cost-saving suggestions and project selection.

Successful model to be applied in other fields as well
One of the challenges was that the track maintenance projects are very heterogeneous, and their attributes evolve constantly. To bridge this gap and to create trust with management, the project focused on rapid prototyping: going rapidly into production and delivering results in from an early stage on. The model provides reduction of the deviation from portfolio cost target by factor 10; the cost driver analysis showed that average costs could be reduced by 4%. Overall, the effectiveness of the projects is expected to increase by 5-8%. The model developed proved so successful, that it will be applied in other project fields at the railway company as well.

From marketing related issues to strategy & planning
As Senior Data Scientist Bojan Škerlak provided an overview over Data Strategy and Data Science at Migros. The complex organization structure of the group with different cooperatives, production firms and companies is a challenge for the retailer. Currently, the overall focus lies on marketing related issues; for example, promotion and targeting. Other examples of cases at the retailer are geo location or the question if and how they can use data from weather forecasts to adjust the offer in the stores and therefore reduce food waste. As federation of cooperatives, Migros also has to work on leveraging synergy potentials systematically. Future focus in Data Science will be strategy & planning as well as supply chain management.

Thanks for your registration!

Machine learning for productionTensorFlow SessionShaping the Data Age Together: SDS 2019AMLD 2022, EPFL LausanneHWZ Yea(h)rbook 2017 FachbeitragMaking NLP easyReal-Time Tracking of Swiss Covid-19 CasesStart up Winji in der NZZD ONE with 3 talks at SDSData Visualization: Snowstorm in AthensSDS2018 - with D ONEArtificial Intelligence in Claim ManagementThe 6th Swiss Conference on Data ScienceVon Daten zum Business Value - Workshop von D ONE und DataikuBuilding a Data Vault using dbtvault with google BigQuery