Blending old world hard metal with future-focused data-driven analytics, Siemens Mobility Data Services division is capitalizing on big data and analytics to ensure transportation around the globe is fast, reliable and more energy efficient. Innovation that includes predicting failures, ensuring a seamless supply chain for parts to reduce or eliminate downtime and use weather data to differentiate problems in the same train model in different regions. WOW!
In 2014, the century and a half old multi-billion Euro company launched Vision 2020, defining a path to a successful future. That vision requires innovation with data and a commitment to the mission, “We make real what matters.”
“I think in a certain way, Siemens has always been data-driven because we have always been engineering driven, and engineers are the ones who really always look at data. They always look at a different set of data. They couldn’t handle large volumes in the past because it was technically not possible, because we didn’t have enough sensors. But they always went through this procedure trying to understand: how can we improve the products we’re making? How do we test them better to make sure that the products are fitting customer needs? Data has always been part of that. Using a different technological basis, not being as large as it is today, but it’s always been at the core of the DNA of Siemens.”Gerhard Kress, Director, Mobility Services
Now that data storage challenges are a thing of the past, Siemens engineers are leveraging tens of thousands of sensors and the data they send. Data from the trains and rails, repair process data, weather data, and data from supply chain all go into Siemens Teradata Unified Data Architecture leveraging Hadoop, Teradata Aster and the Teradata Data Warehouse.
“We could not do what we’re doing based on a different architecture because data volumes we’re having are rather large. So, for example, one fleet of vehicles from Europe, we just gather together all the data from sensors– it was about 100 billion lines of a table. If you want to run a machine learning algorithm on that, that does not work on something that’s not massively parallel. And that’s actually very, very helpful.”Gerhard Kress, Director, Mobility Services
Machine learning with Teradata Aster enables Siemens data scientists and engineers to quickly identify false positives (predicting a failure that doesn’t really happen) and give a clear prediction of actual part failures. Incorporating weather data, Siemens can differentiate what is more likely to fail on the high speed train between Moscow and St. Petersburg in the frigid winter versus the high speed train traveling in the hot Spain summers. Allowing train schedules to be impeccably reliable (Siemens services the Bangkok metro where only 1% of trains are ever delayed!).
Looking at the Spain to Barcelona high speed train – travelers used to fly 80% of the time, now because only 1 in 2000 journeys experience anything more that a five minute delay, only 30% take planes. The train line even guarantees travelers that if their delay is more than 50 minutes they will refund the ticket! That’s confidence and customer service!
“That happens because we have data, we have analytics models, and we can actually predict certain failures. There’s gearboxes, for example, on high-speed trains, it’s one of the things that is most tricky to monitor. We had a couple of cases where you could predict those things would be breaking in a few weeks. We had ample time to provide the spare parts, do the right thing, repair it, take the train out of normal circulation without harming the schedule, and work with the customer without having any problems for them.”
With the vast amount of unstructured data, sensor, telematic (you name it, Siemens has it) Siemens must have an innovative scalable architecture. Gerhard gave an example that from one fleet of vehicles in Europe there were 100 BILLION lines in a table.
“What I actually like is that the Unified Data Architecture is really combined into three different elements, and I believe that’s actually the real strengths. If you only compare it component by component, you might find competitors, but all together as one big architecture, it’s really great. What I also like now is really that it’s moving away from three components being connected on a network to really an integrated platform, a data lake, and that the platform understands where data is, can move it, shuffle it around. With QueryGrid, I don’t need to know where the data is stored. I can just push a query down into Hadoop, and don’t need to load stuff around. And now with the new announcements around Teradata Loom, I believe that’s going to be taking it a step further. So I think the vision is great, and we’ll rely on Teradata implementing that vision.”Gerhard Kress, Director, Mobility Services
Thank you Siemens for sharing your incredible story of success with big data and analytics!