Trains, IoT and analytics: GE’s best kept secret

When we say ‘GE’, you probably think of industry solutions such as power plants; wind turbines, airplane engines. But did you know that GE is also one of the world’s biggest train vendors? Neither did we. Yet, GE has over 17.000 locomotives in operation worldwide. The longest train in operation is over 7 km long, the heaviest weighs over 100 ton.

But, as we learned during SAS’ most recent AX event in Amsterdam, this is not the only impressive aspect of GE’s railroad business. Equally impressive is the use of technology in these giant machines. Most locomotives are, for instance, equipped with sensors that automatically enforce a slowdown of the vehicle when they approach an urban area. 

As we were implementing sensors to enforce some regulatory measures, we might as well use them for other purposes”, that was GE’s reasoning behind the next phase of IoT, explained GE Transportation’s Garrett Fitzgerald GM for Transport Intelligence.

In this next level, edge devices are generating hundreds of sensor data elements per second, providing useful information on the status of each locomotive. By analyzing these data in real time, GE Transportation is able to anticipate on potential problems, and to act on these data even before the incident takes place. Providing the necessary maintenance and replacement of components when required is, after all, a lot cheaper than waiting for an incident to happen and having to provide that maintenance and repair anyway.

The sensors can also contribute to a considerable cost saving in fuel and energy consumption, eg by spotting a running engine on a halted locomotive and turning this engine off. “This may seem trivial”, commented Fitzgerald, “but with an energy bill of 10 to 14 billion dollar per year, this can lead to huge cost-cuttings.

In order to analyze and interpret all these sensor data quickly and correctly, GE Transportation makes use of SAS Event Stream Processing based on SAS Viya. This allows them to analyze sensor data close to the data source, thus saving precious time in gaining valuable insights.

We will also be able to dynamically create groups of assets based on location and their behavior and their status (moving, stationary, ...)”, continued Garrett Fitzgerald, “which will contribute to a more efficient asset management of our distributed and mobile assets. This will again lead to significant cost savings.”

GE’s best kept secret may end up being one of GE’s most valued and inspiring initiatives.