Swiss Railways - flying by iPad via TIBCO

Profile picture for user gonzodaddy By Den Howlett October 15, 2013
Summary:
Optimizing an already dense and complex rail network to squeeze more capacity requires precision. Add in huge amounts of data being generated and you need a highly scalable solution to make it run. Enter TIBCO.

According to Holger Mueller the Swiss Federal Railway (SBB):

That's not quite true. A 2010 case study suggests that SBB was 88 percent punctual within a three minute window whereas most others measure against a five minute window. Under that measure, SBB would be 96 percent punctual. Be that as it may, an early foray into driving efficiency targeted an improvement to 90 percent. This in turn allows SBB to run more traffic on its static infrastructure.

SBB operates a very dense network of railway infrastructure that cannot be physically expanded but which was faced with a need to accommodate additional traffic loads. A central method of overcoming this seemingly intractable situation was to minimize network conflicts. this was achieved through real time messaging and alerts to drivers. These messages, which are generated at a rate of 7,000 per second, are carried on TIBCO infrastructure which in turn feeds business applications that predict traffic speed and movements throughout the network.

A the business end in the driver's cab, speed recommendations are made to coincide with times when the driver can pay attention to iPad delivered information and recommendations. This requires very precise timing.

In the video above, Marcus Voelcker, who was CIO at the time of the original project and is now a consultant for follow on projects talks briefly about the detail involved in this complex process. He then goes on to describe how, having achieved one set of benefits, SBB is now using the system to help work out optimal energy savings within the context of a highly dense traffic flow.

What he doesn't say but which I subsequently learned is that the TIBCO system is managing 1.5 terabytes of data per day. That's huge and gives a perspective on what 'big data' means in the real world.

Show notes:

0:52 - Early stage software platform build: "If you are punctual then you can sell more tracks"

1:50 - Adaptive control: speed control = energy savings

3:02 - GSM sends speed information and recommendation to drivers' iPad

3:24 - Usability is key, understanding constraints is crucial: "No Angry Birds in the cab!"

4:33 - Managing complexity - 1,000 trains in motion per second.

5:50 - Small margins for error

6:07 - What's next? Continued efficiency: "Not fly by wire but fly by iPad."