Wednesday, September 27, 2006

Maglev Train Wreck: The Human Factor

For the past several years, a train that literally floats on air and travels at speeds up to 280 miles per hour has been operating regularly on a 19-mile test track in northwestern Germany. The Transrapid 07 "maglev" train's test runs are open to the public, and the waiting list for a ride often exceeds six months. On Friday, September 22 of this year, some thirty visitors and employees of the train's manufacturers, ThyssenKrupp and Siemens, filed aboard for a high-speed trip along an elevated guideway that wends through forests and pastures. Earlier that day, maintenance personnel had traveled the same route in a smaller service vehicle which normally was moved out of the way to clear the track for the Transrapid. But somehow, that morning the service car was still on the main line when the Transrapid plowed into it at a speed of 125 miles per hour. Twenty-three passengers and crew died and ten more were injured in the most serious accident to befall maglev technology since its inception.

The idea of using a magnetic field to support a vehicle without contacting the ground is not that new. Patents on the basic idea were filed as early as the 1930s, but the notion had to await advances in electrical power and control systems before a practical maglev train could be designed. The first full-scale experimental units were fielded in the 1960s, but so far the only commercial maglev train, a German Transrapid, shuttles between downtown Shanghai and the city's airport.

The technical appeal of magnetic levitation is easy to understand. At train speeds over a hundred miles an hour, stresses on conventional train wheels and tracks become extreme, leading to increased operation and maintenance costs. In operation, the Transrapid makes no physical contact with the track. Instead, powerful magnets hover less than an inch below steel strips on either side of the track, and automatic control systems measure the distance thousands of times every second to keep it within close limits. Heavy copper coils of wire along the track produce moving magnetic fields that propel the train up to 280 mph, eliminating any need to transfer large amounts of electrical energy to the train.

How does it stop? In normal operation, the same moving magnetic fields that accelerate the train also slow it down. The excess mechanical energy that braking makes available can even be captured and sent back into the power grid, making maglev trains one of the most energy-efficient transportation modes around. In emergencies, a mechanical system takes over. The train is fail-safe in the sense that if all power fails on the train and the track, the cars simply settle down on a skid pad on the track and the whole thing just slides to a stop without leaving the rails. All the cars remained on the track even after the recent accident.

So what went wrong? A complete answer must await future investigations, but initial reports indicate that the train operators simply did not know that the service vehicle was still on the track. At speed, any train—maglev, electric, diesel, or steam—takes a long distance to stop, a distance that increases greatly for high-speed trains. Stopping after the driver sees an obstruction is usually not an option. So the whole orientation of train safety since the nineteenth century has been to keep obstructions off the track. And this is largely a matter of good communications between the train operators and those in a position to know what is on the track ahead, out of sight.

A friend of mine belongs to the Austin Steam Train Association, a largely volunteer-staffed organization which operates excursion trains in and around Austin, Texas. Even though what they do is for fun and not for pay, they follow all applicable rules, regulations, and licensing requirements for safe train operation. After years of study, my friend finally got his engineer's "ticket" recently. Even though he is a professor of engineering, he had to undergo a course of study and a rigorous examination about the fine points of train operating procedures, including rules about authorization for train movements that seem almost Byzantine in their complexity. But decades of experience have proven these rules to be necessary, and he takes pride in following them to the letter.

Anyone can make mistakes, and this is not to say that those who operated the Transrapid on that fatal day did not have enough rules and regulations. All the regulations in the world will not prevent an accident if the rules aren't followed, and the fact that the Transrapid operated with a good safety record up till now says that by and large, the operators knew how to run it safely. Perhaps the experimental nature of the maglev train allowed a certain complacency to creep in. Track sensors that detect obstructions and interlock with train controls would have prevented this accident. And perhaps the commercial installation in Shanghai features such safety interlocks. It would be a shame if this mishap, which had nothing to do with the maglev features of the train and everything to do with human error, ends up tainting the future of maglev technology. All the same, it is a reminder that no matter how advanced technology becomes, the people working with it have essential roles to play in making it safe to use.

Sources: A New York Times article describing the Transrapid accident is at Some interesting historical background on maglev technology in Germany can be found at The Austin Steam Train Association's website is


  1. I must disagree with any suggestion that failure of a system that does not have safety interlocks can be described as "human error" unless the human who erred was the designer of the system. Automatic safety interlocks are a standard engineering practice. There is simply no excuse for omitting them and relying on a fallible human system

    Prof. Vincent Brannigan
    Clark School of Engineering
    U of Maryland

  2. Dear Dr. Stephan,

    This is superb reference material for my latest novel. Sincere thanks for posting it.

    David Long

  3. This comment has been removed by the author.