* 23 entries including: railroad and road infrastructure, California road trip, home hybrid power-heating systems, genetically modified mosquitos, prenatal genetic testing controversy, emissions trading controversy, updating the Panama Canal, phone card ripoffs, ICBMs for conventional strike, personal supercomputing, coin hoarding in India, Alpine & Gibraltar tunnels, user interface for XO computer, forensic shoeprint analysis, and the risks of being an archaeologist.
* INFRASTRUCTURE -- RAILROADS (2): In the USA, the standard railroad track "gauge" -- the span across the two rails -- is 1.44 meters (4 feet 8.5 inches). George Stephenson, the English engineer who invented the locomotive back in the 1820s, settled on this gauge, apparently because it was close to the wheelbase of existing horse-drawn trams.
Not everybody accepted this gauge. When the railroads began to take root in the USA in the 1830s and 1840s, gauges ranged from 1.3 meters (4 feet 3 inches) to 1.83 meters (6 feet). The US South standardized on a 1.52 meter (5 foot) gauge -- which actually remains in use in Russia, since the first Russian railroads were built by a Southern American named G.W. Whistler, father of artist James Whistler. In some cases, there were reasons to use a nonstandard gauge, for example to obtain a monopoly on rail traffic in a locale; when rail lines were isolated and disconnected, the variations in gauge didn't make that much difference. However, after the US Civil War, the national railroad system extended its reach, and by 1886 most of the railroad track in America was of standard gauge.
Since trains go through tunnels, under bridges, and pass by each other, height and width of the cars have to be specified as well. In the USA, the maximum height is 4.57 meters (15 feet), while the maximum width is 3.25 meters (10 feet 8 inches). European trains tend to be smaller. Railroads are still not highly standardized -- in Europe, train signals and electric train voltages may vary from country to country.
* Given how heavy trains are, it's a bit surprising that the roadbed for the tracks is so simple -- crushed gravel and crossties. The fact that the roadbed has a certain amount of "give" under such loads is a plus: although it increases friction and makes pulling the loads a bit harder, a more rigid bed would be torn apart in short order.
The ordinary crosstie in the US is wood, and it's a bit surprising for Americans to find out that in Japan and Europe they're usually steel-reinforced concrete. Americans use wood because lumber tends to be cheaper in the USA than elsewhere. Incidentally, in the UK crossties are called "sleepers". A US crosstie is usually made of oak, about 20 centimeters (8 inches) on a side and about 2.5 meters (8 or 9 feet) long. It is soaked with creosote or coal tar, which increases the lifespan from about 5 years to about 30 years or so. Typical crosstie spacing is about 50 centimeters (20 inches). The rails are spiked down to the crossties through steel plates, which generally sit on a rubber or plastic pad. The gravel bed on which the ties sit is called "ballast", for reasons that seem to have been forgotten. The gravel is limestone or harder taprock, crushed up to make sure that it can be packed down into a stable mass.
Track-laying is a highly automated operation, with a variety of obscure-looking machines to do the job. Other machines help maintain the tracks, ensuring that they remain in alignment. One particularly interesting machine used by US railroads is a conventional highway pickup truck with an auxiliary set of railroad wheels that can be raised and lowered -- it is amusing to see a pickup truck speeding down the rails on its auxiliary wheels. [TO BE CONTINUED]
START | PREV | NEXT* GIMMICKS & GADGETS: WIRED.com had a report researcher named David Stone from the Massachusetts Institute Of Technology, who has come up with an interesting, low-cost approach to building simulations: use the Second Life virtual environment as the framework, and use the handheld Wiimote device associated with the Nintendo Wii game system for input. He is promoting such simulations through his WorldWired consulting operation and has a number of clients.
In Second Life, it's relatively easy to build chairs, buildings and other objects for avatars to sit on or walk through. Tools like wrenches or manual controls are also easy to build and, with a little tweaking, users can control them with a Wiimote. The Wiimote is a much more intuitive interface to such simulations than a conventional computer mouse, and it is adaptable -- for example, Stone has built a plastic steering wheel with snap-fit attachments for a Wiimote.
* It always seemed a bit much to scrap a five-year-old personal computer that was generally working right just to upgrade to the latest thing. According to an article on BUSINESSWEEK.com ("Recycled PCs Bridge Digital Divide" by William Sturgeon), now old PCs can be put to good use. Government directives are now zeroing in on the messy business of disposing of old electronic gear, providing incentives to find ways to get rid of PCs in a benign fashion. Computer Aid International is providing a conduit through which old but functional PCs are distributed to third-world countries.
Computer Aid only accepts PCs with a Pentium II processor or better. The hard disk of the computer is completely wiped to ensure that no private data is passed on to end users, and then ensures the PC is up to operational capability. The computer is packed off to undeveloped nations. In Kenya, for example, the PCs are accepted by Computers For Schools Kenya (CFSK) in Nairobi, where they are checked, kitted up with software, and then passed on to schools that are waiting on a list for PCs.
CFSK makes sure that teachers are trained to use the machine and that the support infrastructure for running the PC -- such as electric power -- is in place. To date, about 2,000 teachers have been trained in Kenya, with 7,000 PCs distributed to 300 schools. Course materials are provided for the PCs, which are networked using a distributed system that ensures useful web pages are buffered on local servers, permitting a degree of connectivity even when local datalink technology won't support reasonable access to the full web. Students can share files and send email to each other.
Kenyan students are enthusiastic about the program and are eager learners. Companies doing business in Africa, such as Barclays Bank, are behind the program, Barclays having donated a thousand PCs to CFSK. Some "Greens" have accused the giveaway as just being a "stealth" means of dumping old PCs onto undeveloped countries, but advocates insist that the machines passed on are in good working order and have plenty of useful life left in them.
* BUSINESS WEEK reports that McDonald's provides home delivery in 25 major cities around the world, including Cairo, Taipei, and Shanghai. Scooter drivers deliver the Big Macs and Quarter-Pounders, with a delivery fee improving margins. There's no need to clean up after customers at the outlet, either. There is no way this will happen in the USA, however: the business model only works in really packed cities where drive-through is a troublesome concept, and where labor is cheap.
BACK_TO_TOP* CHP AT HOME: As reported in IEEE SPECTRUM ("A Power Plant For The Home", April 2007), in the US a typical home has a heating system driven by natural gas, and an electrical power distribution system driven by a distant electric power plant. The problems with this scenario are that producing electricity usually generates heat that is simply discarded unused, and pumping electricity over power lines wastes a fair percentage of the power.
The answer in principle is the "combined heat & power (CHP)" system, which provides useful heating while producing electricity. CHP systems are already in use in hospitals, university campuses, large petrochemical factories, and other large installations. CHP is also used for district heating in Denmark, the Netherlands, and other northern European countries. Now "micro-CHP" systems that provide heat and electricity to homes are increasingly popular in Europe and Japan. SenerTec of Schweinfurt, Germany, sells a 5 kW CHP system for apartment buildings in Europe, and Honda has sold 50,000 1 kW systems for home use in Nippon. Climate Energy of Medfield, Massachusetts -- a joint venture of Yankee Scientific of Medfield and ECR International of Utica, New York -- is performing a trial with 25 homes fitted with Honda-built 1.2 kW CHP units. Climate Energy officials expect to sell several hundred units in 2007.
Micro-CHP systems generally consist of an internal combustion engine and a furnace. The engine drives a generator to produce electricity, with the heat produced by the engine dumped into the furnace through a heat exchanger. The engine is usually driven by natural gas, but it can run on other fuels. The micro-CHP system does not end reliance on the power grid, since it only works when the house heating system is turned on; at other times household power is drawn from the power grid. CEO Eric Guyer of Climate Energy say the scheme is straightforward, based on no particularly new or exotic technology, more common-sensible than whizzy -- like a hybrid gasoline-electric automobile, not a fuel-cell car.
A micro-CHP system costs a few thousand dollars more than a traditional gas furnace. It makes the most sense in locales where winters are long and cold, and where electric power is expensive. The US Northeast is a good example, and in fact local utilities in that region are offering incentives to buy micro-CHP systems. In such an environment, a micro-CHP system can pay itself off in two years and return savings of $500 USD a year after that. In more typical climates, it might take ten years to get a payback. It works worst in climates where air-conditioning is more important than heating. A US Environmental Protection Agency estimate claims that 30 million American homes are good candidates for micro-CHP systems.
In principles, excess power produced by a micro-CHP system can be fed back into the power grid, with the system owner being paid off by the local power utility. More than 35 US states now require utilities to provide net metering to track customer power contribution. However, states decide what type of residual energy customers can pump into the grid, and so far only 11 states allow net metering for CHP.
In general, incentives for installing a micro-CHP system are much weaker in the USA than they are elsewhere. In Germany, micro-CHP users are waivered an "eco-tax" normally attached to the use of natural gas, and they are paid more than the wholesale price of electricity for what they pump back into the power grid. Micro-CHP users in the United Kingdom get a big discount on the natural gas tax, paying just 5% instead of 17.5%. In Japan consumers can't sell electricity back to utilities, but the government gives subsidies to support micro-CHP, and gas utilities also give a discounted rate to micro-CHP users.
The US Department of Energy did perform research on a US micro-CHP program in 2004, but then the funding dried up. Climate Energy's Guyer is undiscouraged: "This is the case of the microcomputer versus the mainframe. There were people in the mainframe business who thought that there was no role for personal computing." He adds that a typical gas furnace has a lifetime of 15 to 20 years: "My expectation is that for people who are putting in a new gas furnace now, when it's time to replace that, they'll be putting in a micro-CHP unit."
BACK_TO_TOP* SUPER MOSQUITOS: Malaria is a major problem in the developing world, and one of the traditional tactics to suppress it has been to try to kill off the Anopheles mosquito that transfers it to humans. This is a difficult task on the face of it, which has been getting worse as the mosquitos have acquired increased resistance to pesticides. According to an article in THE ECONOMIST ("A Shift Of Perspective", 24 March 2007), a group of researchers is working on a more Zen approach to the problem: genetically engineer a mosquito that isn't so easily infected with the Plasmodium malaria parasite in the first place. Mosquitos that aren't infected with malaria can't pass it on to humans, helping break the cycle of infection.
Dr. Mauro Marrelli and Dr. Chaoyang Li of the Johns Hopkins University in Baltimore, Maryland, have actually created such a "transgenic" mosquito, using a gene named SM1 that creates a protein to block access of the Plasmodium protozoan through the mosquito's gut. Another approach might be to boost the mosquito's immune system to fight the parasite, but that imposes a penalty on the mosquito if it's not infected with malaria, meaning mosquitos with such a modification would not be able to compete with "wild" mosquitos and would die out. The idea behind the use of the SM1 was that its penalty would be small enough to be more than outweighed by its advantages, allowing the modified mosquitos to eventually outbreed their wild relatives.
To test this idea, the researchers linked in a gene for "green fluorescent protein" along with GM1 to make the eyes of the modified mosquitoes glow green for identification. A lab environment was then created in which transgenic mosquitos were mixed with wild mosquitos. When fed mouse blood infected with malaria, in nine generations about 70% of the population became malaria-resistant.
The experiment was interesting but left open some questions. First, why didn't the transgenic mosquitos drive out the wild mosquitos 100%? One reason appears to be that when mosquitos that have two copies of the SM1, one obtained from each parent, it does impose a significant penalty. (Ironically, the gene that causes sickle-cell anemia in folks of black African descent works in a very similar way: creating malaria resistance when one copy is present, causing big problems when two are present.) The other problem was that the experiment used a strain of Plasmodium that infected mice but not humans, and so wasn't completely valid as a test for human malaria. However, the results were positive enough to suggest that the researchers are onto something -- though the idea of breeding healthier mosquitos still seems contrary on the face of it.
BACK_TO_TOP* CALIFORNIA ROAD TRIP (3): On Wednesday, 16 May, my agenda was to visit the Planes of Fame air museum and Yanks air museum at Chino, to the east of Los Angeles proper. First thing I did was to drop by the Walmart and buy a wi-fi interface. I did get a bit lost trying to backtrack to get out to Chino, teaching me the unwisdom of attempting to navigate by guesswork in a strange city. Fortunately, I managed to get back on track after about ten or fifteen minutes. I learned my lesson: always follow the directions even if it means having to go the long way around sometimes.
Neither the Planes of Fame air museum nor the Yanks air museum are all that impressive as air museums: both have excellent collections, but they don't really have the display facilities to show them off properly. That isn't really a slam, since both operations are basically aircraft restoration facilities -- most of the stuff actually flies, which is very impressive -- and display is something of a secondary objective. Both have serious aircraft boneyards in the back lot. I figured the payoff would be the airshow on Saturday.
I had expected to spend more time at Chino, so I felt I had time to go visit the Long Beach aquarium -- it was low priority, it's hard to get good pix in an aquarium since they're dim-lit and most of the exhibits are behind glass. I snaked through the traffic across town and found the place pleasant. The major interest there was an outdoor cage of lorikeets -- midsized parrots. They would troop kids through them and the lorikeets would land on them to drink out of cups of sweet syrup; I got some good pix of the lorikeets. The QUEEN MARY ocean liner was also on display nearby in the harbor, accompanied by an old Soviet attack submarine. One other interesting item was an offshore oil drilling complex just outside the harbor, unconvincingly camouflaged with fake palm trees.
I went back to my hotel room and installed the wi-fi interface into my laptop. I still have an instinctive dread of installing plugin cards, lingering from my days when I was in Corporate customer service and spent a good portion of my time trying to help people get the wretched things to work. Even ten years ago, plugging a card into a PC was a very dodgy and hit-or-miss proposition. Nowdays, plug it in and it works.
Once I had it working, I then got a bit confused on how to sign into the hotel wi-fi router with the hotel login and password -- it wouldn't swallow them. Huh? I was getting frustrated and then I suspected that I was actually trying to log into the router itself -- to configure it and so on. Obviously hotel guests weren't allowed to do that, so I poked around and then got the guest login screen. Then I was flying. I was entertained to be able to log into my email from my hotel room. [ED / 2021: First time for everything, right? Later email from a hotel room via wi-fi would be expected.]
Another thing that I missed was forgetting that plugging a laptop computer into a power socket in a hotel room can be a pain -- there's sort of a Murphy's law that the sockets in a hotel room will always be in the most inconvenient place, and that they will be loaded up with other appliances. I made a mental note to get one of those cheap multisocket extension cords, a long one, and pack it into my travel kit for future trips. [TO BE CONTINUED]
START | PREV | NEXT* INFRASTRUCTURE -- RAILROADS (1): Chapter 9 of Brian Hayes' INFRASTRUCTURE discusses railroads. Railroads have been around for going on two centuries, and the modern train system retains many features that wouldn't be all that unfamiliar to a 19th-century railroader. A train carries loads on steel wheels on steel rails. This allows a train to carry heavy loads with less friction. The tradeoff of less friction means that trains take a long time to accelerate and a long time to stop. If an engineer sees a disaster looming a mile ahead, there's not much he can do but put on the brakes and, if worst comes to worst, jump off. He certainly can't steer away from the accident -- he can't steer period, all he can do is go forward or backward. Steering the train is handled externally, by switches on the track.
The rails are called "T" rails, since the cross-section looks like an upside-down "T". The base of the rail is called, unsurprisingly, the "base", while the top -- the bearing surface -- is the "head", and the relatively slender component joining the two is the "web". Not all rails are created equal; "high iron" rails that have to support fast and heavy trains are comparatively robust, while the rails of a small branch line may be substantially lighter. The rails are marked to give the date of manufacture, who made the rail, and how heavy it is.
A standard US rail is 11.9 meters (39 feet) long, for the simple reason that rails are generally carried on a railroad car 12.2 meters (40 feet) long. Except for high-speed and high-load main lines, the rails are bolted together from web to web using two "fishplates" on each rail. The fishplates mate to holes in the rail web. The rails are not butted up strongly against each other, to allow for thermal expansion, and that is why trains generally go "clickety-clack". Since communications with trains are traditionally relayed through the tracks themselves, there's generally a copper wire linking two rails across a fishplate. Rail lines are organized as discrete "signal blocks", and so at the boundaries of the signal blocks there are no copper wires, with the fishplates connected through plastic insulating blocks.
On most main-line tracks in the USA, the rails are actually continuously welded in place. The railroads were unwilling to try such a scheme for a long time since it meant that the rail would shift from being under very strong compression to very strong tension as the temperature changed. However, it turned out that the ties and the spikes holding the rails to the ties could handle the forces easily. Welded rail is assembled from the 11.9 meter long segments in a shop into rails some 439 meters (1,440 feet) long. They are carried on special cars, with only about 40 per car to allow them to flex as the train negotiates curves. Once laid, the segments are welded together, often using "thermite" -- a mix of powdered aluminum and iron rust, which burns very hot. The track laying has to be done in middling weather. If it was done in cold weather, the track would only be able to expand, if it was done in hot weather, the track would only be able to contract.
Tracks get rusty fast, so a rail that still has a shiny surface is carrying a fair amount of traffic. Sometimes rails will feature corrugations from braking or turns, and in spots there will be a "locomotive burn", where the engine spun its wheels and dug out the rail slightly. Wheels will get a bit flat as well at times, resulting in a distinctive KABUNK KABUNK KABUNK sound. [TO BE CONTINUED]
START | PREV | NEXT* PGD EXAMINED: An article on MIT TECHNOLOGYREVIEW.com ("Choosing Babies" by Emily Singer) took a closeup on the controversial technology of "pre-implantation genetic diagnosis (PGD)". PGD is a complement to in-vitro fertilization (IVF), giving prospective parents a genetic profile of the zygote that results after fertilization in a test tube. PGD was developed in the 1990s to identify embryos at risk of serious genetic diseases, such as Tay-Sachs disease. Since that time PGD has expanded its domain to cover a wide range of other genetic conditions, many of which are not all that serious.
PGD has been controversial since the beginning, and it's not hard to see why. Suppose a woman with three sons wants to finally have a daughter through IVF, and specifies that only a female embryo is appropriate -- the male embryos will be disposed of. Similarly, suppose a couple with a family history of diabetes specify that only an embryo lacking diabetes susceptibility genes is appropriate -- with all other embryos, once again, disposed of.
With about 50,000 babies conceived via IVF in the USA each year, the issue is far from academic. Parents are now increasingly selecting embryos according to sex, and it's perfectly within imagination that they will be doing for any other trait that can be nailed down -- eye or hair color, for instance. Genetic tests are now available for a thousand conditions, including deadly childhood diseases and adult-onset cancers (in particular breast cancer), and more tests are being discovered every day.
Data on the actual use of PGD in the US is only now starting to become available. A survey of fertility clinics offering IVF performed by researchers at Johns Hopkins University in Baltimore, Maryland, was published in late 2006. The results showed that screening for chromosomal faults that can lead to implantation failure or miscarriage, or for disorders linked to chromosome duplication or deletion -- such as Down's syndrome -- accounted for two-thirds of PGD testing. Tests for genetic diseases such as cystic fibrosis accounted for another 12 percent. Almost half the clinics said they had received requests for PGD that posed ethical questions, with most from parents who wanted to select the sex of a child. In fact, almost one in ten tests was for nonmedical sex selection.
Some fertility clinics will not offer sex selection; in fact, the US is one of the few nations where it is even legal. Some critics feel that it opens the door to other types of nonmedical selection, though others think that's the way of the future, one saying: "I think parents will want to do it, so I think this will expand rapidly."
Even testing for medical conditions has its ethical problems. To be sure, in some cases the tests will reveal unquestionably serious or fatal conditions, but in many other cases all they do is reveal a higher susceptibility. This raises ethical questions, and since more "susceptibility genes" are being found all the time it may eventually be impossible to test an embryo with PGD that doesn't reveal a number of them. Since US fertility clinics are as a rule for-profit operations, they have a bias to maximize profit, instead of ensuring the welfare of clients confused by the results of PGD tests.
In the United Kingdom, a government body licenses fertility labs and regulates which tests can be performed. The US government has left the matter more or less alone, meaning its implementation is left in the hands of the fertility clinics. A certain level of regulation seems inevitable, though some believe that the rights of the individual should come first, one saying that "the final choice, once tests are considered to be scientifically legitimate, should be left up to patients and physicians."
BACK_TO_TOP* FAIR TRADES? There is now a widely-held belief, based on solid scientific evidence, that the threat of global warming from greenhouse gas emissions is real and needs to be addressed. As reported in BUSINESS WEEK ("Another Inconvenient Truth" by Ben Elgin, 26 March 2007), along with legitimate concerns there is a lot of fashionable posturing, in some cases manifested in the form of "green" initiatives that amount to much less than meets the eye -- such as "greenhouse gas offsets" and "renewable energy certificates (RECs)".
The idea behind offsets is that a business or organization that is emitting greenhouse gases can pay off another that is cutting emissions to obtain credits, with the goal being to end up with overall neutral emissions. A REC is along the same lines, adding a premium to purchases of power from traditional sources to help support renewable energy schemes, the idea with the renewable energy "credited" to the purchaser even though the power actually being delivered isn't "green".
Such "trading" schemes have merits and limitations, with the limitations showing up in particular in the USA. Although some governments elsewhere, particularly the European Union, promote schemes like offsets, at the present time the US Federal government does not. As a result, environmental trading in America is pretty much handled any way the participants in a scheme want to handle it. Says one official with an environmental organization: "Right now, it's a no-man's land out there."
One of the big players in offsets in the US is TerraPass of San Francisco, the brainchild of Karl Ulrich, a professor at the Wharton School of Business of the University of Pennsylvania in Philadelphia. In 2004 Ulrich, a dedicated environmentalist, gave his class a few thousand dollars and had his students cook up a workable carbon-offset program. The next year, a group of students went on to found TerraPass, a for-profit company. 30-year-old Tom Arnold, TerraPass's CEO, says that the company now has 42,500 clients and has handled over a hundred thousand tonnes of offsets.
TerraPass takes money from clients and passes it down to businesses that are investing in technology to reduce emissions. One common approach is to reduce emissions of methane, a potent greenhouse gas, by flaring it or burning it to produce power. This produces carbon dioxide in the end, but carbon dioxide is about 20 times less effective as a greenhouse gas than methane, so it's a very good trade from that point of view. One set of projects being backed by TerraPass involve two collectives of dairy farms in Washington State and Minnesota that set up methane digesters to obtain electric power from cow manure.
Sounds good on paper, except for the fact that TerraPass, and a broker who set up the deal, take most of the money. TerraPass sells offset credits for about $9 USD per English ton, but the farmers say they only get about $2 USD of that. Not that the farmers mind, because the methane digesters pay off on their own merits, and the additional revenue provided by TerraPass is perfectly welcome.
Similarly, the Academy of Motion Pictures Arts & Sciences, which awards the "Oscars" every year, also recently handed out "mini-awards" to celebrities who had taken on greenhouse gas emissions. TerraPass offsets were behind it all, with one big component on the other end being a huge landfill run by Waste Management INC (WMI), the trash handling giant, in Springdale, Arkansas. Decaying organic matter in a landfill produces a lot of methane; WMI set up 90 wells in the site to draw off the gas, which is pumped to a central flare to be burned into water and carbon dioxide. However, work on the methane collection system was actually initiated in 2001, partly in response to pressures from the state environmental agency, though the action was voluntary. In other words, WMI would have set up the system whether the company got offset money from TerraPass or not. All the offset money does is tidy up WMI's bottom line. In fact, of the six businesses working with TerraPass to obtain offset payments that were contacted by BUSINESS WEEK reporters, five of them casually admitted they would have made the same improvements whether they got the money or not.
Vail Resorts in Colorado used RECs to make the claim that they were "100% powered by wind", but this is misleading since most of the power is provided by coal, the primary electric power source in the West. The deal was handled by a broker firm in Boulder, Colorado -- which refused to tell BUSINESS WEEK exactly where the REC money was going or how much the RECs cost in the first place. The brokerage did claim their transactions were blessed by the Center for Resource Solutions, a San Francisco nonprofit that monitors such deals. The secretiveness of the deal has still provoked skepticism, with one Colorado environmentalist saying that under the circumstances he had to suspect it was all just "vaporware". Some wind power developers are flatly contemptuous of RECs, saying the money is too trivial to be any motivation to actually develop wind power sites and amounts to no more than "pure corporate marketing and image management."
For real ambition in green trading, it's hard to beat the city of Seattle in Washington State. In 2003 the Seattle City Council initiated a program to make the Seattle City Light carbon-neutral. The initial effort focused on projects such as converting municipal vehicles to a biodiesel blend, but this was expensive and brought limited returns. The easy thing to do was turn to trading.
One of the big sources of offsets for the city of Seattle ended up being a plant run by DuPont in Louisville, Kentucky, that produces refrigerants. The city paid DuPont almost $600,000 USD for offsets at a little under $2 a ton -- a bargain compared to the biodiesel scheme, which cost two orders of magnitude more. However, a DuPont spokeswoman flatly admitted that the company "would have continued with these emissions reductions anyway." To complicate matters further, the state supreme court recently judged that the city had no right to spend taxpayers' money on such schemes in the first place.
Defenders of current schemes say that it is appropriate to reward good behavior no matter what the original motivations were, and outfits like TerraPass are also reviewing their procedures to avoid they won't be perceived as smoke and mirrors. The pressure is rising to make sure the game is played in a more aboveboard fashion, before it ends up being reduced to a complete joke.
BACK_TO_TOP* REBUILDING THE CANAL: The Panama Canal is one of the great engineering wonders of the 20th century, but as reported in POPULAR MECHANICS ("The Panama Canal's Ultimate Upgrade" by Brad Reagan, February 2007), it's in need of some work. The canal is maxed out: it transfers 14,000 vessels, carrying 5% of the world's cargo, every year, operating around the clock at about 90% of its maximum capacity. Traffic jams occur when things go wrong, with canal officials occasionally resorting to auctions to determine who gets to go first after the problems are resolved.
For a long time, ships were designed to fit the canal's 33.5 meter (110 foot) wide locks, but now freighters are being designed beyond the "Panamax" size limits. The Suez Canal, a sea-level cut that can be expanded with relative ease, now handles 20% more traffic than the Panama Canal. Faced with obsolescence, in October 2006 the citizens of Panama voted for a $5.25 billion USD upgrade plan.
The current Panama Canal runs from Balboa on the Pacific coast, just north of Panama City; through locks at Miraflores into Miraflores Lake; through another set of locks at Pedro Miguel into the Galliard / Culebra Cut; through central Gatun Lake; then finally through locks at Gatun to a cut that empties into the Atlantic at Colon. The update project will add two new sets of single-lane, three-step locks, along with two new navigational channels to connect to the locks, while the existing passageways will be deepened and widened. In all, a total of 130 million cubic meters (170 million cubic yards) of rock and earth will be moved, resulting in a new canal that will be able to handle the larger "post-Panamax" vessels -- up to 366 meters (1,200 feet) long -- and double the maximum rate of traffic.
The Panamanian government's Panama Canal Authority (PCA) plans to leverage off digging performed by the USA for an improved canal that began in 1939 and had to be called off in 1941 due to the outbreak of war. The PCA believes that the project can be completed in eight years or less, with under 7,000 workers -- a tenth of the number of workers that built the original canal.
The centerpiece of the work will be the two new lock complexes. Each will be over 2.4 kilometers (1.5 miles) long, with lock chambers 427 meters long, 55 meters wide, and 18.3 meters deep (1,400 x 180 x 60 feet). One of the big challenges in the design isn't obvious until pointed out: water supply. Despite the fact that Panama gets over 250 centimeters (100 inches) of rain a year, and Gatun Lake has a surface area of 430 square kilometers (166 square miles), it is the primary source of all fresh water in Panama, and at present the canal drains off about 7.6 billion liters (2 billion US gallons) of water a day. Adding the new locks threatened to multiply the loss. The solution was to add three catch basins alongside each new lock, which will allow about 60% of the water to be recycled. The new locks will actually use 7% less water than the old locks. The level of Gatun Lake will also be raised about half a meter (1.5 feet) to increase the water supply. In addition, the lake will be dredged out to provide 1.5 meters (5 feet) more draught, and widened from its current 150 meter (500 foot) minimum to 280 meters (920 feet) on straightaways and 365 meters (1,200 feet) on curves.
The new locks will require digging out 8 kilometers (5 miles) of new channels. The old locks use double-leaf doors with a miter seal, pressed shut by water pressure. They work fine despite their age, but when maintenance needs to be performed the lock has to be shut down, blocking traffic. The new locks will use one-piece doors that roll in from the side, but will have twin redundant doors on each end of the lock -- allowing one to be pulled in and serviced while the other keeps the lock operational. The old locks used a set of electric engines known as "mules" to guide ships through, but the new locks will be big enough to allow use of the tugboats that normally guide vessels through the canal.
PCA officials realize the job is a big one, but they think they have the plan and the tools to do it. One says: "If the Americans could build this whole canal in ten years, we can finish this project in seven."
BACK_TO_TOP* CALIFORNIA ROAD TRIP (2): On the morning of Tuesday, 15 May, I got out of Las Vegas early -- noting with no particular pleasure the fine layer of dirt deposited on Ryo Oki by the dusty desert air -- and cruised through the desert regions of southeast California on to Los Angeles. Some folks love the desert, but I didn't have any enthusiasm for it myself -- I found the occasional "cactus forest" alongside the road surreal, like something Lovecraftian. It wasn't just that the cactus were spiny; they grew in amazing twisted-up shapes, as if born deformed.
I entertained myself scoping out power lines during the trip, and there were some particularly elaborate tower farms in that corner of the desert. I was picking out single-phase rural lines, three-phase distribution lines, and trunk lines of various voltages, as cued by what I had read in Brian Hayes' INFRASTRUCTURE. I saw a number of high-voltage lines with dual conductors and a few with triple conductors, but though INFRASTRUCTURE claims that there are trunk lines with four conductors, I never saw any.
I was fearing that LA would be smoggy when I got in and the skies were gray as I entered the north side of town to get to my first objective, the LA zoo. However, the skies then began to clear and what I thought was smog turned out to be just coastal haze. That was the pattern all the time I was there -- gray in the morning, clearing up by midday, with a slight haze lingering. The air rarely appeared particularly dingy.
I did have some problems finding the zoo -- navigating the interstates for two days was no problem, but I ended up squirrel-caging around the zoo area until I figured out the ambiguities of the local direction signs. The zoo is well set up, similar in size and layout to the Denver zoo. I found nothing particularly exotic, the most interesting exhibit being the little delicate gerenuk antelopes, like miniature impalas that liked to stand up on their hind legs to browse on low trees.
After the zoo, I cut across town to pay a visit to Los Angeles airport (LAX). I had a very specific reason for this: I'd picked up many pictures of airliners at Denver International Airport, but I've never seen a Boeing 747 there. I figured that I would see 747s on transoceanic flights at LAX, and planespotting was no problem there -- the end of the runways is separated from the sea by the Dockweiler State Beach, where somebody taking pix of airliners would not attract too much attention. It paid off, too -- I got pix of two 747s on the outbound, which on later inspection turned out to be very high-quality, pretty shots -- one of a Singapore Airlines machine, the other of a Lufthansa ship (and believe it, a 747 is a ship). I also got some pix of LA fire department helicopters in orange-and-white "creamsicle" colors shuttling up and down the shoreline.
Then I drove across town to Anaheim. Although I had been dreading LA traffic, I got to thinking it wasn't so bad: If I could handle Denver traffic, I could handle LA traffic -- the only condition being that LA covers from about four to eight times more surface area, giving it a sort of TWILIGHT ZONE sense of endlessness. There was the usual mix of courtesy and obnoxiousness among drivers -- I don't mind people weaving in and out too much, but would it be so much to ask to use the turn signal? There's every good reason to do so and no sensible reason not to. Ah, I've done my fair share of bungles in freeway driving so I can't get too sanctimonious.
One thing that did startle me a bit was somebody on a motorcycle cruising up the dividing line between the toll lane and the rest of the freeway. I thought: I bet the cops get annoyed with that sort of thing if they catch somebody doing it. No more than moments later a motorcycle cop came passing by in pursuit -- I laughed out loud. Another thing that I eventually noticed was that the posts for highway signs were ringed by razor-wire concertina. I wondered why, and the answer instantly popped into my head: Taggers. I have a particular detestation of vandals -- thievery is nasty but rational; trashing things for the fun of it is beyond the pale.
I stayed at the Anaheim Super 8 -- I usually pick Super 8s when I'm on the road since they're sort of like the McDonald's of motels, cheap, nothing fancy, and I know what I'm going to get. This one was a little seedier than most and the room smelled odd. The last time I had been to Anaheim was in the 1990s and the area was pretty run down, but since then there had been massive investment, with hotel complexes springing up all around. The Super-8 was a survivor from the grungy old days, and given the level of building in the area, I suspect it will either be renovated or replaced in the relatively near future, since otherwise it won't be competitive.
After settling down at the motel, I decided to get online and change my reservation for the trip back home from Las Vegas to Cedar City, Utah. The hotel had free wi-fi access; I didn't have a wi-fi interface for my laptop, but I managed to rent a little time online at an office at the Jolly Roger hotel down the street. Incidentally, walking around in the Disney resort area can be a bit hard on the feet. Sites may only be a few streets away, but the blocks between the streets are the size of parks. With the massive buildup in the area, they do look quite a bit like parks -- some of the interior lanes in the hotel complexes are lovely little "tropical paradises", lined with palms and other greenery.
I felt pretty dim about not been able to use wi-fi access -- I hate to miss a trick -- and while I was changing my reservations I checked on Amazon to price out wi-fi interfaces. They were cheap, I figured I could do as well or better at a Walmart, so when I went back to my hotel I asked the desk clerk if there was a Walmart nearby. Yep, just drive north on I-5 a few miles and turn right on Euclid, there's a shopping mall. It's always a great help on a trip to have a place nearby to shop for small necessities. [TO BE CONTINUED]
START | PREV | NEXT* INFRASTRUCTURE -- ROADS (7): Once upon a time, roads grids were laid out as straight and rectilinear as possible, but as anyone driving through any residential area built in the past few decades knows, the style for streets off the main drag is to make them wind and snake around. The scheme is known as "traffic calming", and it's designed to ensure that drivers don't get into too much of a hurry in the neighborhood, and aren't inclined to drive through if they don't have a reason to be there.
There's all kinds of tricks to achieve traffic calming, from speedbumps to mini-roundabouts to islands and so on. It is of course no longer unusual to block off streets in congested downtown areas and reserve them as pedestrian malls. Emergency services -- firemen and ambulances -- don't much care for traffic calming, pointing out to the citizenry that it slows down the arrival of help in an emergency, but most citizens regard that as the price to be paid.
* Of course, one of the significant features of the road system, at least in big cities, is the traffic jam. Obviously if a truck stalls on a freeway or there's an accident, traffic is going to be backed up, but what is a bit more baffling is that traffic will sometimes "gridlock" for what appears to be no specific reason, and in fact when the capacity of the roadway has by no means been filled. Usually traffic congestion starts when about 20% of the roadway is covered by vehicles, and is guaranteed once it reaches 30%. Such jams can be mathematically modeled as the almost inevitable consequence of an unregulated flow of moving "particles" (vehicles), bunching up and then dispersing. The jam can propagate like a "wave" down the freeway, with vehicles in back crowding up even as the vehicles in front speed out of the jam.
Traffic engineers have come up with schemes to regulate traffic flow and avoid jams. One relatively recent innovation is the "metered input ramp", where a STOP-GO light gates vehicles to enter a freeway one or two vehicles at a time. Using a metered input ramp system can feel vaguely like taking off an aircraft carrier -- getting the green light and then accelerating as fast as possible to get into the traffic flow.
* The "technology of the future" for dealing with heavy traffic is the fully automated vehicle. The idea goes back decades; early on, it was just a pipe dream, but the latest high-end cars have fancy systems such as cruise control, warning radars, GPS receivers, lots of built-in computer processing, and even head-up displays. Adding sensors, communications, and control gear to roadways is expensive but not anywhere near as expensive as building new roads.
However, although the technology doesn't feel like it's out of the question any more, it's still a big jump to take. In something that's as dangerous as a car, few are going to be eager to hand over control to a computer -- as critics point out, computers are well-known for crashing. Over the short term, it may be preferable for drivers to have "advisory" systems that tell them what the optimum speed and route should be, diverting them to alternate routes in case of accidents. Since human drivers are prone to error, sometimes deadly ones, there will likely come a day when computers will do the job better than humans can, and then most folks will be happy to just tell the car where to go and sit back for the ride -- but when that will happen is anybody's guess. [END OF SET 8]
START | PREV | NEXT* PHONE CARD RIPOFF: BUSINESS WEEK magazine seems to enjoy printing articles on scams and ripoffs, and an article in the 23 July 2007 issue ("Talk Isn't So Cheap On A Phone Card" by Brian Grow) set its sights on popular prepaid phone card.
The phone card is the archetypical gift card, and it is particularly popular with migrant workers in the USA -- in fact, revenues run to over $4 billion USD a year. Getting a permanent phone hookup in America means having a social security number, and those who don't have such a thing have no alternative but to buy phone cards. That wouldn't be so bad in itself, except that some of the phone card vendors don't deliver what they promise. Says a Mexican national working at a fast food joint in Atlanta: "The number of minutes the card tells you that you have, you get half that. It's what everybody thinks."
The trick is that the vendors tend to add various auxiliary fees that are explained in the fine print, if they're explained at all: a connection fee, a disconnection fee, a long-talking fee, and so on. The matter came to a head in March 2007 when IDT, the biggest phone-card vendor, sued nine competitors in Federal court for fraudulent practices. The irony was that IDT also uses auxiliary fees, and some of the defendants in the lawsuit claim that IDT is just playing a game of legal intimidation to step on competitors.
Major retailers such as Walmart also sell phone cards, but they're generally regarded as being on the level. As far as the sleazy operators go, only 11 US states regulate phone cards, with the rest relying on general consumer-protection laws -- and enforcement has been lax. Says a victim: "They're taking advantage of the situation of immigrants. If you don't get your minutes, who are you going to complain to?" The US Federal Trade Commission has been getting an increasing number of complaints, however, and it seems likely that stronger laws and enforcement are coming down the road.
BACK_TO_TOP* REACH OUT & TOUCH SOMEONE: According to AVIATION WEEK, the US Air Force (USAF) had been floating ideas about converting Minuteman intercontinental ballistic missiles (ICBMs) for conventional strike, able to hit any target in the world in less than an hour after the decision to engage. This notion has now congealed into a contract to Northrop Grumman, which will develop a weapon system based on an Orbital Sciences Minotaur booster, which consists of the first two stages of a surplus Minuteman II ICBM lofting two stages derived from the Orbital Pegasus booster. The Minotaur will loft a warhead carrying a cluster of BLU-108B/B Sensor Fuzed Weapons (SFWs), a weapon currently in service as an air-launched cluster munition.
Each SFW looks like a short length of pipe. After dispersal, an SFW pops out a small parachute to make sure it is falling straight down and fires two small rockets that set it spinning rapidly. The SFW then pops out four pucklike "skeet" submunitions that are slung away in four directions. The skeets are unguided but have a sensor to pick up targets beneath them; the sensor is smart enough to determine the class of target and fire its "explosively formed projectile" warhead for shotgun, penetrating rod, or cannonball effect.
The US Navy, not to be left behind, is also working on developing a precision-strike conventional warhead for the Trident II D5 submarine-launched ballistic missile (SLBM). The conventional Trident warhead will have four submunitions. Each of the US Navy's fourteen OHIO-class ballistic missile submarines can carry 24 missiles; the plan is to reserve two missile tubes for conventional Tridents and carry nuclear-armed Tridents in the rest.
There is no commitment to fielding such systems yet. Congress is a bit nervous about ICBMs and SLBMs armed with conventional warheads, since a launch might be misinterpreted as a nuclear strike. However, the likelihood of performing a launch of more than one such missile at a time is very slight, and advocates believe that schemes can be set up to ensure that those who might worry about such a launch, such as the Russians, get proper notification. Besides, much the same concern exists over "quick response" satellite launches, and the Air Force is working hard to making them a reality.
Neither the Air Force nor Navy scheme provides a "hard target" capability, with penetrating warheads that could destroy caves or deeply-buried bunkers. Plans are being drawn up for a next-generation global strike capability that could perform such attacks.
BACK_TO_TOP* PERSONAL SUPERCOMPUTERS? Modern personal computers have a massive jump in capability from the days of the Apple II. According to an article on MIT TECHNOLOGYREVIEW.com ("The Promise Of Personal Supercomputers" by Kate Greene), future personal computers may well be an even greater jump ahead of the personal computers of today -- if they can be made to work.
Intel Corporation has now announced a single chip containing 80 processors or "cores", with the chip able to crunch numbers at a rate of a trillion floating-point operations a second -- a "teraflop". The first computers with teraflop capability were introduced a decade ago and were big mainframes. Intel is evaluating the technology for commercial introduction in 5 to 10 years.
Building a "massively multicore" chip itself has its challenges, but the real challenge is making the chip useful. Simply because a chip has, say, ten cores doesn't mean it can run software ten times faster than a chip with one core: not only does the multicore chip have to access the same "shared resources" -- memory, mass storage, and so on -- but breaking down software tasks so they can be efficiently run in "parallel" is a tricky problem. Intel engineers believe it can be done, and that parallel processing will be particular important to the "recognition, mining, and synthesis (RMS)" applications, such as realtime language translation, realtime video search from voice or image input, and smarter "recommendation systems" such as those used by Amazon.com to recommend products.
Processor performance has been increasing for decades, with clock speeds running at ever higher rates. However, a few years back the writing was clearly on the wall that clock speeds weren't going to increase much more, since the power consumption of the processors was becoming unacceptable. The logical answer, from the point of view of silicon, was to go to multicore chips. However, that immediately raised the issue of parallel programming.
It's not that parallel programming is impossible; it's routinely done on supercomputers for research purposes. The difficulty is that the problems for which it is done are those which can be sensibly broken down into parallel operations. In some cases, this is simple: a multicore chip can be used to hunt for a cipher key by assigning a range of possible keys to each core. Weather simulations are more complicated, with the atmosphere cut up into boxy "cells" that have trade parameters with each other at the boundaries; each core can crunch a specific cell, but they have to exchange data with each other after each set of calculations.
Unfortunately, the sorts of tasks generally run on a PC are not always so easily broken down for parallel processing. It may be possible to get them to work in a parallel fashion, but it may be difficult to do it in such a way where the overhead doesn't make it more bother than it's worth. Parallel programs are also notoriously hard to debug.
Still, special-purpose massively multicore chips are now on the market: graphics company Nvidia is selling a graphics processing chip with 128 cores to provide high-resolution realtime graphics for video games and simulations. Progress is being made on the design of algorithms for parallel processing, with Intel, AMD, and others collaborating with academia to create a common "framework" for parallel processing. One of the academic researchers believes such efforts are likely to obtain results: "In academia, we can disagree for years, but industry has a way, with its economic imperative, to settle on a solution pretty quick."
* AVIATION WEEK had an interesting related article on a new chip built by Raytheon, the "MOrphable Networked micro-ARCHitecture (MONARCH)" polymorphic processor, for military-aerospace applications. Monarch is built around a "field programmable computer array (FPCA)", built up of:
The FPCA is intended for digital signal processing applications, such as radar; it can be reconfigured on the fly to perform different tasks. The FPCA is supported by six on-chip "reduced instruction set computer (RISC)" processors and 128 megabytes of RAM.
BACK_TO_TOP* CALIFORNIA ROAD TRIP (1): I was thinking over for a while on where I wanted to go for my usual spring road trip, and finally decided I'd attend the "Planes of Fame" airshow in Chino, California, in the Los Angeles area, on Saturday, 19 May 2007. That led to deciding that I'd do the full South California tourist thing -- stay at a hotel near Disneyland, since it's set up for tourists and basically a safe part of town, and see the sights in LA and San Diego.
It was a little too long to drive in one day, so I figured I'd do a night layover in Las Vegas -- last time I was there, it seemed like a cheap place to stay, and I could get some nice photos anyway. So I got online and started making reservations, printing out maps and road directions -- even got a very useful satellite picture of Disneyland. I was a bit apprehensive going to South California -- LA's got a reputation for smog and really bad traffic jams, and I could see the trip being more hassle than it was worth. Worse, they'd been having wildfires in LA and I could see the smog as being intolerable.
* I took off in my still basically new Toyota Yaris, Ryo-Oki (that's what the plates say) at 5 AM on Monday, the 14th -- I didn't have a really long day's drive to Vegas, but I had to drive south from Loveland on Interstate 25 into Denver to link to Interstate 70 going west over the Rockies. If I left later I'd get caught up in Denver rush-hour traffic and I didn't want the hassle and delay. All went smoothly through Denver and turned west to cruise up I-70 towards Vail. There was a lot of construction on the eastern slopes of the mountains, with the scale suggesting a lot of money involved -- I saw some mansions on the high ground that were the size of hotels, absolutely mind-boggling.
The region became sparser in every regard as I came back down the western slopes, turning gradually from mountains into the incredibly rugged badlands of eastern Utah, in the region around Moab -- a country of buttes and mesas, layered terrain and wind-sculpted stones. It looked like good country for buzzards and rattlesnakes and not much else. However, as I cruised into central Utah to link from I-70 to turn south on I-15, the land turned much greener, becoming cultivated farmlands and hills carpeted with small bushy trees. The contrast was all the more striking because in southern Utah, south of Cedar City, I-25 winds through sagebrush and cactuslands, cutting across a corner of Arizona though unreal canyonlands to end up in Nevada.
There's a town named Mesquite on the state border, and it's sort of a Vegas in miniature. I was getting in a hurry, however, and didn't investigate, finally arriving in Vegas by late afternoon. That gave me some time to prowl around.
As it turned out, it gave me all the time I needed. I'd been there before and knew it the place was kitschy, and the town tended towards the littered and dirty off the Strip where the casinos were, but this time around it only seemed more kitschy and even more dirty. My recollections that things were unusually cheap also proved unreliable, and I found the entire in-your-face ambiance irritating. It was warm and surprisingly sticky-humid for a desert environment.
I got some interesting pictures -- one of the Excalibur casino, a fake fairy-tale castle, gave me the notion that it had been put together by a megalomaniac miniature golf course designer on recreational drugs -- but though I had planned to stay in Vegas on the way back home, I decided to reconsider that idea. [TO BE CONTINUED]
NEXT* INFRASTRUCTURE -- ROADS (6): The modern red-yellow-green traffic light was derived from a similar warning light used by the railroads (though not using exactly the same color codes). A traffic light assembly is called a "signal head", while the subset of lights facing one direction is logically called a "face", each individual light is called an "optical unit", and the colored lens over each optical unit is a "roundel". The signal head used to be made of iron or steel, but these days it's almost always made of aluminum or high-impact polycarbonate plastics. Head colors used to be black or dark green, but an industrial yellow -- formally "federal yellow" -- color is coming into wider use.
Optical units come in two sizes, either 20 or 30 centimeters (8 or 12 inches). They used to be lit up by incandescent bulbs, but now they generally use arrays of light-emitting diodes (LEDs) -- and in fact traffic signals were one of the first applications for use of LEDs in illumination instead of as simple indicators. LEDs are more expensive than incandescent bulbs, but they last far longer, and if one goes out the light still remains generally functional. LEDs are getting steadily cheaper in any case.
There's usually a controller box on a street corner to switch the traffic lights. Before the days of digital electronics, controllers were based on a motor turning disks that activated relays, switching lights on and off. The old systems generally operated by a fixed, inflexible sequence, with changes in the sequence requiring hardware adjustments to the control system.
Digital control systems were introduced in the 1970s, with a standard specification published in 1992. Modern traffic light control systems can be very flexible. A major intersection might have up to 20 signal faces and 50 optical units, with the control signal factoring in button presses by pedestrians and, using electromagnetic induction loops buried under the road, the presence or absence of vehicles in specific lanes. Since things can always go wrong, the system includes "fail-safes" -- for example, a hardwired system that can tell if current is flowing to green lights in two directions at once, triggering a fault mode.
The induction loop is not the only kind of sensor used to detect vehicles -- ultrasonic, infrared, and other schemes have been used as well -- but it's the most common. Single loops are the most popular, but sometimes "figure-8" loops are used to provide greater discrimination, to pick up motorbikes and even some bicycles, as well as to prevent "false identification" of vehicles in adjoining lanes. Optical sensors are also often mounted on traffic lights to pick up flashing light signals from emergency vehicles, and in some places buses, to then automatically turn the light green. Video cameras are increasingly mounted at intersections to monitor traffic as well.
Traffic lights are often supervised from a central control station. It's not a new idea, with Manhattan operating such a system as early as 1929, but it's more sophisticated now. It's not unusual to see antennas for wireless relay systems on traffic light poles. Traffic light systems can be adjusted to ensure appropriate traffic flow at different times of the day. New technology is also being developed for traffic light systems. Israeli researchers have been conducting experiments in which the controller system uses the sensors on the approaches to the intersection to determine if a collision is likely, and then flashes a warning to drivers to avoid a collision. Like almost everything else these days, traffic lights just keep getting smarter and smarter. [TO BE CONTINUED]
START | PREV | NEXT* A QUESTION OF VALUE: The idea of making mass-circulation coins out of precious metals has long gone out of fashion, but "precious" is something of a relative concept. As reported in an article from BBC.com ("Sharp Practice Of Melting Coins" by Subir Bhaumik), in India the face value of low-denomination coins is less than that of the metal they are made from. The result is that millions of Indian coins are smuggled into Bangladesh to be converted into razor blades and other items -- resulting in a severe shortage of coins in many parts of India.
Police in Calcutta recently arrested a grocer who was operating a coin-melting unit in an unobtrusive run-down shanty. The grocer confessed to melting down tens of thousands of coins, adding: "Our one rupee coin is in fact worth 35 rupees, because we make five to seven blades out of them. Bangladeshi smugglers take delivery of the blades at regular intervals." Police say that originally the raw coins were smuggled into Bangladesh for processing, but the operation has become more sophisticated, with the coins melted down locally for smuggling.
The coin shortage is so bad in places, particularly Indian states close to Bangladesh, that some tea gardens in the northeastern state of Assam have made their own "small change" out of cardboard for internal use by employees. The cardboard coins are the same size as the real ones and their value is marked on them. Says a manager of a tea garden, who insisted on remaining anonymous: "We will commit an offense if these cardboard slips go out, but we have to use them in our gardens because there are hardly any Indian coins in circulation here."
Intelligence officials of the Indian revenue service say millions of Indian coins are being drained into Bangladesh, and have asked the paramilitary Border Security Force (BSF) to be on the lookout for coin smugglers. The Reserve Bank of India, the country's central bank, has been issuing coins energetically to make up for the coin shortfall, but it's like pounding sand down a rathole -- metals traders simply pay a premium for the coins and melt them down. Shopkeepers try to encourage customers to buy more to get a round figure on the price in hopes of not handing out scarce change. They will hand out toffees and other small items as an alternative to change if necessary.
[ED: During the US Civil War, Northerners were inclined to hoard coinage, so postage stamps were used as small change. This was formally endorsed by the government by the printing of stamps without the stickum, which then led to little bills for denominations from 3 to 50 cents. It would seem that this might be a solution for the Indian coinage problem, but I suppose a careful analysis of benefits and drawbacks might reveal some problems. I think ultimately the solution will be electronic cash. It might be cheaper for governments to issue simple electronic-cash gadgets to poor citizens than continue to try to mint coins and bills. Incidentally, handing out toffees is not all that surprising -- penny candy has often been used as small change in countries where coinage is in short supply.]
BACK_TO_TOP* TUNNEL THROUGH THE ALPS: Tunnels tend to be impressive engineering feats. According to an article on BBC.com ("Swiss Dig World's Longest Tunnel" by Imogen Foulkes), the Swiss are working on a particularly impressive tunnel to link Zurich and Milan.
There are already tunnels through the Alps, such as the well-known Mont Blanc Tunnel, but they are heavily loaded and, since they are relatively short tunnels fed by winding mountain roads, tend to be chokepoints. Freight traffic through the tunnels has doubled over the last decade, making the congestion worse. In 1994, the Swiss voted in a referendum to mandate that all freight cross their country on railways. This implied development of railway links to carry the freight.
The key is the Zurich-Milan rail link, which will include what will be in completion the world's longest tunnel, some 57 kilometers (35 miles) in length. The reason the tunnel must be so long is that the link is designed to carry high-speed trains, running at 240 KPH (150 MPH). That would reduce the transit time from Zurich to Milan from the current four hours to two-and-a-half hours, faster than flying when airport time delays are considered. However, it also means that the route has to be as straight and level as possible, with the line maintaining a constant altitude of about 500 meters (1,650 feet) above sea level.
There are actually three separate sets of tunnels involved in the project. The rail line will originate at Milan; run to the relatively short Ceneri Tunnel; emerge into the light for a distance; dive into the long Gotthard Tunnel; run aboveground again for a distance to the Zimmerberg Tunnel; and terminate at Zurich.
Work on the Gotthard Tunnel currently involves 2,000 people, working around the clock, to bore twin railway tubes and access tunnels. Huge fans pump in fresh air, but it's still sweltering in the tunnel, since it's buried under 2.5 kilometers (1.55 miles) of mountain. The workers cut their way in using explosives and the world's biggest tunnel-boring machine, 10 meters (33 feet) in diameter, capable of chewing through 40 meters (131 feet) a day under good conditions. The pressure of the rock above the tunnel is so great that steel rings have to be inserted to keep the tunnel from collapsing.
Surprisingly, problems arise not because the rock is too hard but because it is, in places, too soft, like butter, complicating the digging and slowing it to a crawl. The problems have caused the price tag to almost double, and the tunnel is not expected to be finished before 2018.
Local communities remain generally enthusiastic about the project, though there are concerns that it will marginalize towns that depend on tourists, with the scenic tour through the mountains replaced by a fast dash through a hole in the ground. However, below the town of Sedrun, a 1,000-meter (3,280-foot) elevator shaft was dug to haul workers in and out of the tunnel work site. The elevator, the world's tallest, will be retained in the operational tunnel, linked to an underground rail station, the "Porta Alpina (Gateway to the Alps)". Not only will the underground station provide access to the scenic wonders of the Alps, it may well become a tourist attraction in itself.
* In related news, an article on BBC.com ("Africa & Europe Set For Tunnel Link" by Richard Hamilton) says the governments of Spain and Morocco are now in discussion towards plans to put a railway tunnel beneath the Straits of Gibraltar.
Studies are being conducted to examine the sea bed under the straits to see how difficult the digging would be, with this effort to be completed before the end of 2007. The chief technical expert involved at present, Swiss engineer Giovanni Lombardi, has worked on big tunnel projects in the past, including the Mont Blanc Tunnel in Switzerland, but he sees the Gibraltar Tunnel as an entirely unprecedented challenge: "No works in the world compare to this one. There are a lot of challenges. First of all, the sea at this point is 300 meters [1,000 feet] deep -- about five to six times deeper than the [English] Channel Tunnel. Then there is the geological conditions. There are quite a lot of tectonic movements between the African and the European plates. So there would be quite a lot of movements in the earth, of stresses and so on."
Like the Channel Tunnel, the Gibraltar Tunnel is envisioned as consisting of twin railway tunnels with a central service / emergency tunnel. Along with passengers and freight, the trains will carry automobiles. Government officials on both sides of the strait claim they are now fully committed to the tunnel, with work to begin in 2008. Along with the economic benefits, the tunnel is seen as a historic symbol of will.
Moroccan and Spanish citizens have reservations. The Channel Tunnel has been plagued by cost overruns and financial difficulties, and one Moroccan student commented: "I wonder if the problem of illegal immigration will be solved or not. By the time the tunnel is built, young people will see the tunnel as a chance to run away." However, politicians and engineers believe the tunnel is both practical and desireable. It will not, however, be cheap, with the cost estimated as running from the equivalent of $8 billion to $13 billion USD. Two publicly-owned companies, one in Spain and the other in Morocco, are now trying to raise the funds, and looking for funding assistance from the European Union.
BACK_TO_TOP* XO INTERFACE: The XO, the "$100 PC" being pushed by Nicholas Negroponte of the Massachusetts Institute of Technology, has been discussed here earlier this year; according to a BUSINESSWEEK.com article ("The Face Of The $100 Laptop" by Steve Hamm), along with its other innovations, the XO also offers an innovative user interface or "shell", developed by software designer Chris Blizzard and named "Sugar".
While the XO uses a version of Linux as the operating system, the development team felt that an off-the-shelf user interface wouldn't do. The XO's target audience is mostly children in undeveloped countries, and the standard "desktop" shell used on other PCs simply didn't seem like a good fit.
Sugar has its own distinctive look and feel. When booted, Sugar displays the "XO" logo, stylized to look like a person with arms upraised and called the "XO Man". A dark border frames the display, containing icons to invoke an email utility, a text editor, a photo program (to exploit XO's built-in camera), a web browser, an instant messenger utility, and a web browser.
There are also icons representing the XO's three different operating modes: "Home", "Friends", and "Neighborhood". They provide different levels of interaction with the PC and other PCs networked to it. The low-power XO "mesh network" connects from PC to PC; any PC in the mesh that has access to the world internet can act as a gateway for the others.
In "Home" mode, clicking on an icon launches a particular activity on the PC itself. In "Friends" mode, icons are displayed showing friends operating on the network, along with icons showing what activities the friends are up to; if several friends are collaborating on a particular activity, they are clustered around the icon for that activity. The "Neighborhood" mode gives an overall map of the individuals and clusters of friends on the network and the activities they're involved in. A "zoom" feature allows changing level of focus.
The XO software system is being developed by a small core of professionals, working with the open-source software community. The system uses simple icons and short text labels, along with an interface that encourages exploration. Students using the XO can click and drag pictures found on the internet to a "drop-off" spot or "pocket" where they are stored for handling later. The "Journal" gives the users something like a "blog" to record and share their experiences online. Once a Journal reaches a certain size, earlier entries are automatically buffered on a more powerful computer on the network. Documents and photographs can be stored in the same way.
Applications are written in Python, a simple language that the more technically-minded students can easily learn themselves to write their own applications or modify existing ones. Sugar is only now entering formal usability tests, but preliminary informal trials show that kids simply dive in and stay involved for hours.
* The bit about Python caught my eye. I've been needing a simple but modern programming language to support my work. I tried Javascript, but it's really not designed as a general-purpose tool. The older versions of MS Visual BASIC were just about ideal, but the current incarnations are professional tools -- the download of the freebie version was over 200 MB -- and gross overkill for the jobs I have in mind.
It turns out that Python is easily available online and I promptly downloaded a copy of it -- the whole package is maybe 10 megabytes, an easy download. Tinkerings with it shows it to be easy to use and potentially very powerful. Certainly if it's the default programming language for the XO it's going to get a lot more exposure. Hopefully that won't result in turning it into a cumbersome monster as happened to Visual BASIC. I would think not -- the language is modularly extensible and enhancement seems to be mainly a matter of adding modules as desired.
I was curious as to why the language was named Python, but on obtaining a starter book on the subject, all was made clear: Python's inventor, a Dutchman named Guido van Rossum, is Monty Python fan. Python programming books usually feature examples citing Monty Python scripts:
if weight["woman"] == weight["duck"] then:
witch = TRUE
print "BURN HER! BURN HER!"
The page that contains accessory modules for Python is called the "Cheese Shop" (no, it isn't empty), and the name of the interactive development environment was massaged from the expected acronym IDE into the acronym IDLE. Rossum has, incidentally, been declared "Benevolent Dictator For Life (BDFL)" for Python. [ED / 2022: Been fiddling around with Python on and off ever since. I'll get my notes finally done in 2022.]
BACK_TO_TOP* THE MAKING OF THE FITTEST (17): As a footnote to the discussion of evolutionary genetics, the somewhat confusing concept of "noncoding DNA" needs to be considered in more detail.
The fundamental assumption of genetics after the discovery of DNA was that genes produce proteins -- "one gene one protein" was the (not completely hard and fast) rule. It came as a bit of a surprise that only a few percent of the human genome was actually able to code proteins. What was the rest of it doing? Was noncoding DNA just "junk"?
While some noncoding DNA is known to have regulatory functions, some of it really is clearly junk -- broken genes have already been mentioned. Some viruses can also genome into the genome of the host cell they infect, and sometimes they splice it into the genome of host germ cells. Any creature that is born from those spliced cells carries the virus as part of its normal genome and can actually come down with the disease without being exposed to the virus. Humans have retroviral genes in their genome, but fortunately they are all broken genes, no longer able to be expressed. Such true junk genes or "pseudogenes" do have a useful function for researchers at least, since as discussed in earlier installments they allow tracing the genealogy of species.
What about the rest? What is particularly puzzling is that there doesn't seem to be any particular correlation between genome size and the taxonomy of organisms -- some organisms that seem to be of similar complexity have genomes of very different size. Why does an ordinary onion have a genome five times bigger than a human genome? And why do members of the same genus as the onion have genomes that vary by a factor of four in size?
In many cases, some noncoding DNA sequences can be trimmed out of an organism's genome and the organism will be none the worse for it, suggesting that if those sequences do anything, it's nothing very important. However, some noncoding DNA sequences are conserved, being relatively unchanging over time, suggesting they might actually be doing something, since they're subject to selection. In fact, there does seem to be a clear correlation between the size of the genome of a species and the size of the cells in that species, and the size of the cell does seem to have selective advantages or disadvantages, depending on circumstances. Maybe it's more like finding a piece of wadded-up newspaper in a package and thinking it means something -- when it's only function is as padding. We're not in a position just yet to make many claims about noncoding DNA, and have to regard it for the moment as simply raw material for future Nobel Prizes. [END OF SERIES]
START | PREV* INFRASTRUCTURE -- ROADS (5): There's an elaborate infrastructure of traffic signs and signals that we tend to take for granted, though in fact they're the product of long effort and experience. The first traffic signs in America were set up by bicycle clubs in the 1890s, with automobile clubs picking up the effort in the following decades. State governments got involved in the 1920s. The first international standards for traffic signs were issued by a League of Nations committee in 1931.
In the USA, the stop sign is the only octagonal sign, while yield signs are inverted triangles and warning signs are diamonds. Red is reserved for imperative messages -- STOP, DO NOT ENTER, WRONG WAY, and so on -- while yellow is used for warnings, and orange is used for construction and maintenance signs. Green highway signs provide directions; brown signs indicate parks and other sites of interest; blue signs indicate filling stations, food, and lodging. A fluorescent yellow-green has been introduced for pedestrian-crossing signs. Illuminated signs are usually just silkscreened designs on an aluminum panel. These days, non-illuminated signs usually have a prismatic background to reflect headlight beams back to a driver.
It used to be that all lines and markings on a road surface were just painted on, but now they are more generally high-durability extruded plastic, sometimes with reflective beads mixed in. The little reflectors often embedded into roadways are called "Botts dots", after their inventor, Elbert Dysart Botts. They provide much better feedback under wet or light-snow conditions than painted markings. They usually have white or amber reflectors, with blue dots indicating fire hydrants; some have a red rear face to give drivers the warning that they're going up an exit ramp or the like.
Trying to attach the dots to a roadway proved a challenge, since if they were anchored by bolts or spikes, that would present a hazard to traffic if the dot was knocked off. They are attached by an adhesive. They are more common in warm climates than cold ones, because they don't stand up well to snowplows. [TO BE CONTINUED]
START | PREV | NEXT* SHOEPRINTS: In an era where the public is saturated with the CRIME SCENE INVESTIGATION (CSI) series of TV programs, it's easy to stereotype the work of CSI as focusing on DNA and chemical analysis, but as reported by an article on BBC.com ("Treading Carefully To Fight Crime" by Paula Dear), forensic investigators still pay plenty of attention to shoeprints. Shoeprints are the second most common type of evidence found at crime sites. An imprint of a shoeprint can reveal not merely the brand of a shoe but the particular pattern of scuffs and damage unique to a particular shoe. In the UK, shoeprints can now be taken from suspects who have been taken into custody even if they haven't been charged with a crime.
Now the UK has acquired a new resource for shoeprint analysis in the form of the "Footwear Intelligence Technology (FIT)" system, set up by the UK Forensic Science Service (FSS) to provide data to the nation's police forces. One component is a database of 13,000 shoeprint patterns to match to shoeprints taken from a crime site. Another component is a search facility that will link shoeprints and other data to known criminals.
Jonathan Goodyear of the FSS's footwear section believes shoeprints are a very useful investigative tool: "We may have 10 scenes with the same footprint, but that does not tell us who the offender is. But if at one of those scenes there is a fingerprint, or DNA has been left behind, that would give us a name. It is an intelligence tool to actually link crimes together."
Shoeprints are usually not conclusive in themselves, but they can provide supporting evidence. For example, four regional police forces in England collaborated with forensics specialists to deal with a series of aggravated burglaries, in which the criminals smashed down doors with a battering ram and then performed armed robberies on the victims. A set of 25 shoeprints linked the crime sites and a number of the suspects; seven men were convicted and six ended up with a total of 74 years behind bars.
A shoeprint expert can examine a print for pattern, size, wear, and damage features. Prints can be obtained even when they can't be seen by the naked eye by using powder and special lighting. However, it's not a cut-and-dried procedure either, particularly because a particular shoe changes over time. There is also the problem that two different suspects may have the same brand of shoe and shoe size.
Civil libertarians worry that shoeprints can be used to push through miscarriages of justice. Goodyear responds: "What you need are enough features, such as damage features, to satisfy a scientist." Simply matching a shoeprint to FIT wouldn't be valid evidence: "To get that evidence we would need to seize the shoe, do a detailed examination of it, comparing it with the crime scene mark, and then present that evidence to the court."
[ED: Somehow this reminds me of the corny old cartoon in which Holmes and Watson are inspecting a set of shoeprints on the floor of a crime scene, with the shoeprints neatly marked out and numbered: "I say, Holmes -- how did you ever deduce that the culprit was a dance instructor?"]
BACK_TO_TOP* OCCUPATIONAL HAZARDS: That absolute source of the true facts, THE ONION website, recently recycled an article from 2002 that seems worth placing here, concerning Pennsylvania State University archaeologist Edward Whitson.
Whitson's most recent field trip, to Hasake, Syria, involved the examination of a late Bronze Age settlement, with his primary objective being to determine whether the artifacts there were Persian or Assyrian in origin. Instead, he ended in flight from sphinx demons as he jumped from ledge to ledge in the cavernous halls of a buried temple. Says Whitson: "I was actually only after ancient pottery shards. I didn't plan on being chased by yet another hell-born manifestation of an ancient deity's wrath."
Over the course of his 27-year career, Whitson has earned a towering academic reputation in his field for his willingness to go the extra mile in pursuit of archaeological treasures -- but he's paid for it in his terrifying encounters with pits of giant rats, chimeras with the heads of women and bodies of snakes, and vengeful angels of death. Whitson says that he's learned from his experiences: "I've figured out that I shouldn't read the inscriptions on ancient door seals out loud, and not dust off medallions set into strange sarcophagi. And for replacing jewels that have fallen from the foreheads of ancient frog-deity statues -- well, that's just bad archaeological practice."
Still, Whitson hasn't been able to completely reduce the risks of his work: "Just once -- just once! -- I'd like to be able to perform an excavation without having to listen to the ear-splitting shrieks of the undead while being attacked by glowing scorpions the size of lobsters." He adds: "Yes, I understand that I am often entering sacred grounds, but that doesn't mean I need to be pelted with poisoned darts or have to be trapped between spiked walls slowly sliding together. A simple DO NOT ENTER sign in hieroglyphics would suffice."
To make matters more frustrating, the encounters have little scientific value: "It might be interesting to obtain insights into Etruscan culture from the spirit of a long dead god-king, but no, it's always: I WILL DEVOUR YOUR SOUL! -- or: I WILL RIP THE FLESH FROM YOUR BONES!"
Whitson now feels like he's had his fill of field work: "I couldn't count the number of times I've had to prevent the Apocalypse. I'm too old for this sort of thing, and I'm hunting for employment opportunities as a museum curator. It's unfortunate, nothing compares with the thrill of an archaeological discovery, but after the last time I was forced to swing on a vine over a pit of burning lava with zombie warrior ghouls on my tail, I finally had to decide that enough is enough."
[ED: As usual, I heavily rewrote the article. I kept having the "Indiana Jones March" going on in my head -- only to hear Professor Whitson shout: "Will you STOP that stupid MUSIC?!"]
BACK_TO_TOP