Helsinki Data Science Meetup September 2018 - Analytics goes on the Train

Welcome to the third Helsinki Data Science Meetup of 2018! This time the meetup will take place hosted by VR on Thursday the 20th September, 2018.

Our location is at the VR Group Headquarters located on Radiokatu 3 in Pasila, Helsinki.

The VR Group is a Finnish State owned enterprise that primarily operates in Finland, but also has operations abroad, including in Russia and Sweden. The VR group employs 7,500 professionals and prides itself on providing its customers with high-quality, environmentally-friendly passenger and logistics services. The VR Group has three main business operations revolving around customer groups. VR’s passenger services offer public transport services in commuter and long distance trains and buses. VR Transport offers road and railway logistics services, while VR Track is focused on infrastructure, maintenance and supplies railway materials. In September, we will hear about and learn from their success stories.

The meetup agenda is as follows:
17:30 Snacks and Networking
18:00 Welcome words and our data science architecture in AWS
Annika Nordbo, Head of Data Science, VR Passenger Traffic
18:10 Finding sales anomalies with causal impact

End-to-end solution to detect local anomalies (music festivals, sport events etc.) Data from unrelated train connections was used as reference time series and the actual modeling was done with Google's CausalImpact. Model is ran daily on AWS EC2, results are stored to Snowflake DWH and visualized with PowerBI.
Passenger count predictions

We are using XGBoost to predict the number of passengers in each train at a given point in time. One of the most important parts of our architecture is measuring the training-serving skew from the multiple different models in production.
Heikki Pulkkinen, Data Scientist, VR Passenger Traffic

18:55 Break
19:10 From laser scanning of wheels to simulation of rolling dynamics: An example of condition based maintenance in practise
We have implemented a simulation algorithm that is currently being used to monitor wheelset dynamics. Maintenance decisions based on this data have significant improvement on wheelset lifetime and maintenance cost of SR2 locomotives.
Tuomas Karavirta, Data Scientist, , VR Maintenance
19:50 Mingling