The Right Track to Zero Downtime
In May, my 10-year-old got to experience the bright lights, sights and sounds of New York City for the first time. The printed train ticket was his birthday present. We had planned to depart from Washington, D.C., but at the last minute decided to stay north of Philly and catch an early train out of 30th Street Station to maximize our long weekend.
Three days later, and a little more than a week before we were scheduled to leave, Amtrak made headlines when locomotive 601 derailed at Frankford junction just after leaving 30th Street Station. I was taken aback by the coincidence and concerned for the passengers, but I was also worried that we might have to cancel our trip—the line was now closed for extensive repair.
Trains have a disadvantage as a source of transportation because there usually isn’t a good way to reroute them in the event of a major system disturbance. One solution to overcome failure and delays on the line would be to have another, redundant set of tracks running parallel along the entirety of the same route. But that would be extraordinarily costly and it would still create significant delays; the train would have to return to the station or a junction to get back on course.
All of this thinking about trains and waiting for news about the line at 30th Street Station got me thinking about the we way transport data here at GE, and especially about our PROFINET high availability “line.” Our high availability solution is a redundant control system that keeps critical infrastructure on course through disturbances—keeping hospitals, data centers, power generation stations and even trains up and running 24 hours a day, 365 days a year.
Redundancy in control systems is not new; however old-school systems rely on dual bus topology. Much like a train track, the only way these systems can overcome failure is to employ a completely separate set of “tracks”; two duplicate parallel networks are needed to achieve a reliable backup connection to the central “station” of a network. This type of duplication also requires additional switches and media converters, which add significant cost with little return.
GE’s high availability control solution relies on a ring topology and PROFINET. So if there is a failure or delay on the line, the data can simply be rerouted to where it needs to go in less than three milliseconds (3 ms). There is no need for parallel paths when your system can swiftly and efficiently determine an alternate route without rerouting to switches and converters. This type of media redundancy protocol provides each node on the network with a backup physical connection to every other node on the network, but in a much more cost-effective and direct way: a ring topology. If you’re interested to understand how this works, check out this white board video created by our friends at PROFINET, or download the white paper.
As for our trip, we made it; the trains in the Philly area came back on line with a few days to spare and we had a great (and unforgettable) weekend adventure; although next time we might plan to fly or drive instead…