‘Big data in railway operations is not the future, but a day-to-day business’
Professor Stefan Marschnig from the Graz Technical University has summarized the lessons learned at the Intelligent Rail Summit on 22, 23 and 24 November 2016 in Naples (Italy). Below you can find his speech.
Ladies and Gentlemen,
I have the task and the pleasure to summarize the 3 days of this year’s summit.
I think after these days we are even more sure about that the digital world and big data is not the future but a day-to-day business. We heard about installed devices providing reliable condition data every day helping us to understand better what is going on with our assets. We also learned about what we can expect from new prototypes and on-going developments in future. On the third day we could experience that handling the enormous amount of data is challenging but possible in order to generate the information we need.
I think the railway sector is well prepared for the challenges of the digitalisation era. Of course, we have issues to be worked on in future. I would like to point out two aspects – having been well addressed also by our speakers:
1) Segmentation of the Railway
We face a very complex system of suppliers, infrastructure managers, railway undertakings, and maintenance contractors, to name only some of the main stakeholders and remember what Marc Antoni from UIC said about complex, un-manageable systems. But his is reality and we can’t turn back time. On the one hand we know that digitalisation asks for open data transfer and open interfaces. The railway sector needs to take up speed in sharing data, sharing information in order to be competitive with other transport modes.
It was addressed several times on all three days that we could generate much more benefit in sharing data within the system. We have to overcome especially the legal barriers and create trust.
2) There is the risk that we generate much too much data just because
a) It is cheap to install sensors and store data
b) We cannot solve point 1), so different stakeholders measure the same things.
In my eyes, it is central to understand the data we measure and challenge them with our engineering models and knowledge. “We don’t need analyses telling us that cant and curve radius show a strong correlation, hopefully they do!”
And it is also necessary to combine as much existing data as possible – we learned that today from Prof. Galar and saw it in a practice example from Peter Juel Jensen yesterday. Currently, we still have problems – even in an organisational closed system, the infrastructure – to extract and link data from different sub-systems.
To summarise that, we have to speed up! Not in terms of technology, it often seems there that the speed is even too high to follow it in implementation. But speed up in terms of cooperation and a common understanding of what are the necessities of the Railway 4.0.
Thank you again for being here joining the conference. I hope the mixture of theoretical approach and practical examples, of users, suppliers and academics set up by the programme managers Matthias Landgraf, Markus Enzi and Andreas Schöbel met your expectations.
And I hope to see the one or the other of you at one of the upcoming venues – we keep you informed.
Professor Stefan Marschnig
Technical University of Graz
Are you interested in Digitalisation in Railways? Please have a look at the conference programme of Day 2 during the RailTech Europe 2017 Conference at 29 March in the Netherlands: https://www.railtech.com/railtech-2017/conference/digitalization-in-railways/