Interconnected infrastructure is vulnerable

Last weekend a power blackout completely paralyzed Hamburg's airport.  And last year, a computer glitch also caused chaos.

It's far from being the only place where such things happen: In December, an electric fire caused chaos in Atlanta, at one of America's largest Airports.  The collapse of the flight schedules had knock-on effects throughout the United States and beyond, because connecting flights are so closely interwoven and turnover times at airports so short that one delay leads to another.

Large communities — complex systems

It is not just in air travel that large systems can collapse. In December 2017, when a new train time table took effect in Germany, there were massive delays. The reason was the introduction of the new European Train Control System (ECTS), which is supposed to facilitate the control of high speed trains throughout Europe with all its different technical standards.

What made the glitch even more embarrassing to German rail operator Deutsche Bahn was that most delays occurred on a newly opened high speed line, connecting Munich and Berlin.

But even old and proven grids are not immune to breakdowns. In 2015 a switch tower caught fire in Mühlheim, paralyzing train traffic in a large part of Germany's heavily trafficked industrial Rhine/Ruhr region. In such a case, the train trouble quickly spreads to the roads and the usual rush-hour traffic jam is made a lot worse.

Why is high-tech not foolproof?

It's legitimate to ask why our highly developed industrial countries are so vulnerable to such paralyzing system errors.

In principle, one would assume that all systems have built-in redundancy and undergo preventive maintenance. If, for example, there is a power outage, an uninterruptable power supply would take over until generators start working as a backup. This would ensure that all essential systems, including computers, continue operating.

The same goes for the computers themselves. Servers of large companies are always doubled. If one half of the system breaks down, the other takes over automatically. In most cases, this happens without users even noticing.

Ideally such redundant systems are not even located in the same place — at different branches of a company or, preferably, in other cities or countries. If one building burns down, the data is still secure in the second place. Or, if there is a power outage in one town, the company website can still be accessed from the other location.

Redundancy is not everything – interconnectivity creates dependencies

The truth is: Redundancy protects us from a lot of system breakdowns. If the hardware architecture of most companies were not as well designed as it is, we would experience problems like at the Hamburg Airport on a daily basis.

But at the same time, the systems have to cope with increasingly complex tasks, because technology takes over more and more of our work. That's why a seemingly minor incident can have devastating follow-on consequences down the road.

Low complexity, simple solutions

In the case of a small airport in a remote country, power outages may be part of the daily routine. That's why people there can handle them better: The airport may only be open during daytime, so pilots can fly on sight. If the baggage conveyor belt doesn't work, passengers may be requested to collect their luggage at the plane. If the computers don't work, airport officials issue preprinted boarding passes and have travelers appear on a prepared passenger list.

But that's not possible at an airport the size of Hamburg, which deals with 18 million passengers per year. Despite all built-in redundancy, situations can lead to overload. And simple pragmatic solutions may be out of the question given high-priority security concerns.

In such circumstances, it would be enough for the baggage conveyor belts to be out of service, or the scanner at the gate not functioning properly. If, as it happened in Hamburg, the power outage takes numerous smaller sub-systems out of service, an orderly operation may not be possible anymore even though the sub-systems themselves are not essential to the whole operation.

High complexity, big capacity

The more rarely such an incident occurs, the greater the risk that a system reaches its limits when it happens, because people have less experience and routine in dealing with such a situation.

If we expect trains to travel at speeds of 300 kilometers (186 miles) an hour, we must rely on highly complex control technology. If a system breakdown forces the trains to go at much lower speeds then of course this will mess up many timetables, not just that for one specific route.

But there is some good news too: After all, such breakdowns in highly complex systems usually get fixed within a matter of hours or days. And once things are back up and running, it takes only a day or two for the systems to reach their full capacity again. And all this is thanks to — not despite — the complex technologies we so much depend on.