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 http://www.nytimes.com/2006/09/23/world/europe/23cnd-germany.html?_r=1&oref=slogin. Some interesting historical background on maglev technology in Germany can be found at http://maglev.de/index.php?en_vision. The Austin Steam Train Association's website is www.austinsteamtrain.org.

Tuesday, September 19, 2006

Email: Boon or Bane?

If you are reading this blog, you must be on the Internet, unless you are standing outside my office door where I post a hard copy of the first page every week. Either way, you very likely have one or more email accounts. If you're like me, your feelings about email are not uniformly positive. Sure, it's convenient, cheap, and a great way to stay in touch with people on the other side of the world, if you happen to know anybody over there. But email's downsides are well known too, from the time it takes you to wade through spam up to career-destroying incidents that could have been prevented by thinking just a little longer before clicking the "send" button.

Like most other communications media, at first glance email doesn't seem to have much to do with engineering ethics. A saying around my house is, "More communication is better than less communication," and why wouldn't that apply to email? The first large-scale users of email were physicists who found it a convenient way to keep in touch via their advanced networked computers. Many of the standard features of email were developed in an environment where the users were intelligent, well-behaved, technically adept, and often had a libertarian streak that opposed excessive government regulation. The protocols and systemic features that email uses were developed in this environment. Consequently, email's marginal cost is basically zero, anyone with an email account can send mail to anyone else, and it is almost impossible to regulate without extensive government-funded intervention, as in China.

These features stayed in place as the volume of email grew far beyond what most of its early developers anticipated. Now it is a part of modern culture, as much as the telephone was half a century ago. The near-zero marginal cost of email has allowed spammers of all kinds to spring up, and a kind of electronic warfare now exists between spammers who spray the Internet with billions of bits of advertising in the hopes that a few people respond, and the system operators who keep improving spam filters in a constant battle to limit the junk percentage in the average user's in-box. One wonders how much resources are being wasted on both sides. If there was a tiny fixed cost to sending an email message built into the system, even a Federal tax, and that cost was impossible to avoid, most spammers would go out of business. The rest would have to behave more like direct-mail companies, carefully targeting their messages to only those persons who are more likely to respond, given the limited financial resources of the spammer. The horse has been out of the barn much too long to consider implementing something like that now, unless in a few years it becomes necessary to do a worldwide system upgrade reaching down to the very basics of the email protocols. And that is not likely for the foreseeable future.

Spam aside, even the volume of email from people and organizations I recognize is often overwhelming. I find that the two largest generators of emails I realize are legitimate, but which I'd rather not receive, come from the two universities I am associated with, one as an employee and one as an adjunct professor. To control this problem, there would have to be some kind of financial or other penalty associated with excessive use of the all-employee email list. Most organizations have some sort of policy along those lines, but its enforcement is sporadic and sometimes you wonder if anyone cares at all how many emails are sent out to everyone.

Finally, there is the time each individual spends dealing with email. I must personally spend an hour or more each day dealing with it: reading it, sorting it, purging it, filing it, writing responses, and so on. In the years before email, what did I do with that hour or so a day? I don't remember reading postal mail for that length of time daily. And I wasn't on the phone. I must have been able to do other useful things, such as work, reading important books, or talking with friends and relatives. Whatever it was, it doesn't happen now, or if it does, it's in the rest of the day that has been squeezed by email.

Email isn't the first communications medium that has been viewed with ambivalence. Plato, writing around 380 B. C., called into question the wisdom of the invention of writing itself. In a story he attributed to a legendary king, he noted that before the invention of writing, people had to commit important things to memory: songs, poems, even legal agreements. But now that the technology of writing was available, the skills of remembering would atrophy. He was not at all sure that writing was an unmixed blessing.

Neither am I sure that email is an unmixed blessing. But I hope that in the future, its rough edges get smoothed out and it approaches the ideal of a seamless meeting of minds that all communication should strive for. It hasn't happened yet.

Sources: The Socratic dialogue in which Plato recounts the encounter between the Egyptian divinity and inventor of writing Theuth and King Thamos is available at http://english.ttu.edu/kairos/2.1/features/brent/platowri.htm (Phaedrus 67-71).

Tuesday, September 12, 2006

Death in Africa for Cell Phones in the U. S.

According to some estimates, four to ten million people have died in the war that has raged in the Democratic Republic of Congo since 1996. Africa's third-largest country was known as Zaire until 1997, and began its sad history of relations with the West as the Congo Free State in 1870, when King Leopold of Belgium made it his personal property. The despicable exploitation and cruelty that Leopold wrought upon millions of Africans in his efforts to extract natural resources such as rubber and diamonds reduced the population by half in thirty years, and has ever since stood as a paradigm of human rights abuse. Today, the Congo holds another material that the rest of the world covets: colombo-tantalite ore, commonly known as "coltan." And although there is no single individual like King Leopold who can be held responsible, the Congo is once again suffering horribly as the rest of the world steals its treasures.

Coltan is the world's main source of tantalum, an essential element in the manufacture of miniature electrolytic capacitors, also called "pinhead" capacitors because of their size. Without these capacitors, portable electronic equipment such as cell phones, PDAs, and iPods would either be much larger or simply wouldn't work at all. When only expensive military gear used tantalum capacitors, the demand for coltan was small. But now that consumer electronics manufacturers use millions of them, coltan is a hot commodity in the world mineral market.

The U. S. has no significant natural deposits of coltan. Other than Australia, the largest reserves are in the Congo. Makers of consumer electronics buy tantalum capacitors whose ingredients very likely come from a country where illegal mining, smuggling, and full-scale warfare over coltan-rich regions is endemic. The detailed history of the Congo and coltan is complex and tangled, involving multinational companies in the U. S. and Europe, migrations of refugees from Rwanda, interference by the government of Uganda, and general bad behavior on all sides. (For more information, see the article by Keith Harmon Snow and David Bernouski at http://zmagsite.zmag.org/JulAug2006/snowpr0706.html.) But the simple fact is that much of the coltan that makes its way into the world's supply chains of electronic components was mined either illegally or under political or moral conditions that most people would be horrified at if they knew.

So what is an electronics engineer to do? Avoid any designs that use tantalum capacitors? That's hardly practical, and for one thing, you can't tell just by looking whether the tantalum in a particular device came from Australia, the Congo, or somewhere else. But if engineers simply shrug their shoulders and say, "The supply chain isn't my problem—if a part's price is right and meets the specs, I've done my duty," then the professionals who are in the best position to know about the situation and make decisions based on it are turning their backs on the problem.

In Europe, certain activist groups have publicized the connection between portable electronics and the murderous events in the Congo, chanting "no blood on my cell phone" and calling for an embargo on tantalum from illegal mining. But embargoes and boycotts are not as effective as professionals who organize to recognize a problem and take action against it. At the very least, those who specify and use components whose ingredients may have been extracted at the cost of human suffering should be aware of the sordid background behind some innocent-looking electronics parts. And if their consciences moves them to do something about it, so much the better.

Engineers and technology specialists should learn from the food industry, where product differentiation has been raised to a high art. Most consumers can't tell organic broccoli from the other kind simply by tasting it. The U. S. Department of Agriculture has developed a "certified organic" system which tells the consumer that organic produce was grown without pesticides and so on. What the consumer pays extra for is not necessarily a better taste, but the knowledge that his vegetables were grown in a certain way. Some makers of clothing feature the fact that their products were not made under sweatshop conditions that prevail in certain parts of the world. Again, the intrinsic quality of the goods is not in question. What is being sold is a feeling or sense that the purchase is somehow making the world a better place.

Why can't this principle be applied to consumer electronics? First, an auditing system of supply chains would have to be implemented so that one could trace supplies of raw materials all the way back to the mine. Given the corrupt nature of some governments and institutions, this would be hard. But if certain firms managed it somehow and made enough of a big deal with publicity and advertisements, the fickle hand of the consumer might begin to favor the firms taking such trouble over those who were not making sure that their products did not use materials that contributed to human exploitation. It sounds silly and idealistic, maybe. In 1820, the idea of banning slavery in the U. S. sounded silly, idealistic, and dangerous. But those who believed in it persisted, and now, slavery is virtually unheard of, at least in the West.

Here is a challenge that goes beyond engineering, but needs engineers and other technologists to implement. The only thing that is lacking is the will on the part of those involved to do something.

Sources: Besides the Snow and Bernouski article noted above, Snow has many other articles on exploitation of African nations by multinational corporations at his website http://www.allthingspass.com. The boycott efforts are described briefly in a BBC article at http://news.bbc.co.uk/2/hi/africa/1468772.stm.

Tuesday, September 05, 2006

Does Engineering Ethics Change With Time? The Case of the 1955 Oldsmobile

Owing to circumstances too complex to summarize here, the other weekend I found myself working on the electrical system of a 1955 Oldsmobile Super 88 sedan. The original owner needed help in getting it running again after it had sat unused in a garage for three years. All it needed was a new battery, a set of ignition cables and distributor parts, and a little gasoline poured down the carburetor to give the gas pump a chance to suck new fuel from the gas tank. After a few tries, it started up fine and made all of us very happy, just as it did in 1955 when it became the owner's first new car.

While working around the vehicle, I was struck by the tremendous differences in design and engineering between what was considered a good, safe, responsible-family-man's car in 1955, and what would pass muster by today's standards. The biggest difference is the list of what is missing from the '55 compared to what you would find in virtually any new car today. That list includes: air bags, computers, shoulder belts, high-impact bumpers, a catalytic converter, numerous other anti-pollution technologies, anti-skid brakes, steel-belted radial tubeless tires, and unibody construction with added side-collision protection. On the other hand, what the '55 has in abundance that has been greatly reduced in most modern cars is expressed in one word: steel. The main longitudinal members of the chassis under the car look like they're strong enough to support a skyscraper. The bumpers are thick chrome-plated armorplate that must weigh close to a hundred pounds each. Carry this theme through the whole car, and you have a behemoth that needs a 300-cubic-inch V-8 just to get moving. And of course, the gas mileage (premium only, now that leaded gas is no longer generally available) can't be much better than 20 mpg.

Because you couldn't legally market a new car designed like this 1955 specimen today, does that mean that the engineers who designed it then were bad engineers? That they were doing something wrong in foisting such a dangerous vehicle upon the public? Of course not. No one is blamed for not using safety technology that doesn't yet exist. Today's air bags, for instance, depend on a micro-electromechanical IC to detect the rapid deceleration of a collision. Anti-skid brakes need computer technology for similar reasons, and such advances were simply not available in 1955. Seat belts, on the other hand, are hardly advanced technology. According to a document on the Ford Motor Company's website, 1955 was the year Ford became the first U. S. automaker to offer seat belts as an option. But it took decades for car companies to offer seat belts and later air bags as standard equipment. And until states began to pass laws mandating the use of seat belts around 1980, only about one out of ten drivers used them.

There are two morals to this little history lesson. The first one is pretty obvious: standards of what is considered safe and legal design change with time. Laws change as a matter of course, and engineers need to be aware of any changes in laws that affect their firm's products or services. But always staying just barely legal is not usually the best position for either a company or an individual. The best engineering looks ahead of the present legal environment to combine what new technology offers with desirable features that make products safer as well as better. And if you widen the concept of safety to include broader public goods such as less environmental pollution or less fossil fuel consumed for a given output, the range of potential engineering development grows even wider.

The second moral is related to the first: even if laws and technology change, changes in behavior don't always follow. It took a combination of public-awareness campaigns and legal sanctions to get most (not all!) drivers to buckle up. Many today would view the emission of leaded-gas fumes from a 20-mpg 1955 vehicle as at least as dangerous as the absence of seat belts. Many cars sold today get no better mileage than this, and the main reason is not technology, but buyer preference.

There is a debate going on today about how we in the U. S. can reduce our dependence on foreign oil imports. Some favor increasing mandatory federal fleet mileage standards, but auto manufacturers have shown how they can "game" that system in the past. Others take the view that as gasoline becomes more expensive, Americans will just naturally start buying more fuel-efficient cars. A third answer is to radically increase the taxes on gasoline to more accurately reflect the "externalities" of driving gas-hog vehicles: the cost of importing oil, the cost military measures to stabilize the Middle East, the hidden costs caused by air pollution, and the highway-construction costs associated with more cars on the roads, period.

While the means vary, the goal is the same: to change the way drivers behave toward driving and cars. And changes in behavior usually come slowly, if at all. But given the right combination of suitable new technology, economic incentives, and what for lack of a better word can be called the cultural factor, people can change the way they behave pretty fast. With most drivers having cell phones, it's trivially easy to call 911 when you see a road accident, and help can arrive much faster than when someone had to find a pay phone and call the local authorities. If the desire to help others wasn't there, all the cell phones in the world wouldn't do any good. But before it was so easy, the Good Samaritan who would undertake either to drive for help or to stop and render aid was not that common.

That sort of thing was probably not in the minds of cell-phone designers, but it is an unexpected benefit. Engineers can intentionally look for such benefits and design them into new projects, but only if they take a wider view of the ways their work will be used and consider the human factors as well as the purely technical ones.

We still haven't taken the '55 Olds out on the road. The brakes need a little work yet. But when we do, it will only be for special occasions, perhaps car shows, where we can show younger generations what dangers their parents faced in the bad old days of primitive technology. And will today's 20-somethings in their Honda Civics look adventurous and risk-taking, or profligate and careless about the environment, to their grandchildren? Only time will tell.

Sources: The Ford press release describing their introduction of safety belts in 1955 is at http://media.ford.com/newsroom/release_display.cfm?release=23485. The statistics on belt use are from http://en.wikipedia.org/wiki/Click_it_or_ticket.