The risks of telecom

The October 22 2012 issue of The New Yorker has an interesting story (http://www.newyorker.com/reporting/2012/10/22/121022fa_fact_osnos) about corruption in China with the July 23 2011 train crash outside Wenzhou.

One detail caught my eye. The story described how the railroad had to use a backup system of wired boxes by the tracks because the radios (the articles said radar) didn't work in tunnels – boxes which failed in this case. Yet at the very same time that the train engineer was unable to reach the dispatcher, there was a passenger who was not only able to, but actually uploading a picture to the web! It seems that the railroad is dependent upon a traditional telecom model limiting communicating to one pre-designated path. With the Internet model it doesn't matter what path the bits take – if someone could upload a video then one should be able to communicate with the train.

This is one of the many consequences of path-dependency in creating (unnecessary) failures. Admittedly I have idealized this the story a bit because today's Internet is still implemented in terms of paths thanks to the business model of broadband (and cellular) but what is important is the larger idea in which the relationships are not tied to any particular path and thus can be resilient. In China there are extensive cellular networks – why aren't they available as infrastructure for the railroads? 
Shared publiclyView activity