Misjudged

 

On October 22, 1895, the train you see in the photo left Granville on the Normandy coast at a quarter to nine in the morning with destination Paris. As the crow flies, that is about three hundred kilometers (186 miles), which takes a modern train three and a half hours. At the time, the journey took an entire day.

At 3:55 p.m., the train rumbled into Paris, but it was several minutes late. The very experienced driver, Guillaume-Marie Pellerin , thought that he could partly make up for this delay by braking only at the last moment. But this time, the brakes failed and the train crashed through the buffer and glass facade of Paris-Montparnasse station, where it came to a stop, as the photo shows, in an unreal position. There was one fatality to regret. Marie Augustine Aguillard wasn't even a passenger on this train – no, she was minding her husband's kiosk at the Place de Rennes for a while; he had gone to get the evening papers. She was killed by falling debris.

The American George Westinghouse had invented a brake based on compressed air some twenty-five years earlier. The brake engages when the air lines are deflated and will not release until a compressor has repressurized the lines. Because each carriage has its own brake, the entire train is braked. That system seems inherently safe: if something breaks, the pressure drops and the brakes kick in. However, on this train the Westinghouse brake failed anyway, and the brakes of the locomotive alone could not stop the train in time.

Engineer Pellerin took a risk. Has he thought carefully about what could go wrong and what the consequences could be – precisely at this location, a terminus? His train's inherently safe brakes gave Pellerin enough confidence to brake a little later than usual. If he had looked just a little further, he might have thought that if the brakes failed, it could have disastrous consequences in this very spot.

Risk is often expressed with the simple formula Risk = Likelihood x Severity. We often do not calculate with numbers, but with estimates: small, medium, large – both left and right possibly flanked by 'very'. The formula shows that an event, which is unlikely to occur (Westinghouse brake failure), can nevertheless lead to a high risk, because the expected consequences are very serious (deaths and injuries). The limits of the risks you want to take are determined by your risk appetite. Adventurous people have a greater risk appetite than cautious people, and manufacturers of hip technology products take greater risks than a government organization, just to name a few extremes.

You yourself also perform risk analyses every day, for example when you cross the road. You make an assessment of whether you will make it before that car reaches you, and you mainly look at the distance and speed of the car, and how well you are on your feet. But do you also consider the possibility of tripping? Do you still have enough time to get away, or does the driver have sufficient reaction time and is his braking distance long enough? We usually don't think about such a scenario, probably because it usually goes well. And that was precisely Pellerin's problem. It cost him a fine of fifty francs and two months of suspended prison.

Do me a favor and take care crossing the street when you go out later.

There will be no Security (b)log next week.

 

And in the big bad world…

This section contains a selection of news articles I came across in the past week. Because the original version of this blog post is aimed at readers in the Netherlands, it contains some links to articles in Dutch. Where no language is indicated, the article is in English.

• the FBI hacked Russian state hackers.
• this security company is very open about an attack on their systems.
• the British plead for more transparency in cyber attacks.
• Rian van Rijbroek's husband seems to be quite confused himself.  [DUTCH]
• you should watch out for fake windows updates. 
• you should also watch out for malicious QR codes.
• the emergence of AI and the possibility of cloning on the fly may lead to additional legal protection for individuals.  [DUTCH]
• YouTube is experimenting with banning ad blockers.
• many Dutch people can protect themselves better than they do now.  [DUTCH]
• phishing is the main form of cybercrime that citizens have to deal with.  [DUTCH]