Thoughts on business, engineering and higher education

Today in History, From the Engineering Pathway

Here are links to recent posts from the "Today in [Engineering] History" blog that I have particularly enjoyed:

• June 5, 1977: Apple II goes on sale.
• July 22, 1975: Bill Gates and Paul Allen license Basic.
• September 19, 1982: the Smiley emoticon is invented :-) Can you imagine a world without Smiley? That post really is a must-read.
• September 30, 1935: Hoover Dam is dedicated. 
• October 1, 1958: NASA begins operations.

I absolutely love the "How Everyday Things Are Made" website hosted by Stanford's Innovative Alliance for Manufacturing, which has a lot of informative videos and was featured in the "National Manufacturing Week" post by Alice Agogino. Make sure you watch the video on how jelly beans are made.

Also, since Google bothered to change its regular image to a barcode today because the barcode was patented 57 years ago (October 7, 1952), I'll mention that I wrote a post a while back about the scanning of the first barcode in June 1974 for the Engineering Pathway, post that was updated today in honor of the event. Thanks to Alice for pointing it out. A blog post in the Christian Science Monitor mentions the lesser-known fact that the barcode patented in 1952 consisted of a series of concentric circles. To see for yourself how ugly that looked, click here.

The Economist's New Columns

The Economist started two new columns in recent months: Banyan, about Asia, in April, and Schumpeter, about business, last month. This is in addition to Lexington about the US, Charlemagne about Europe, Bagehot about Britain, Buttonwood about finance, as well as the Economics focus column, also in the finance section, which will hopefully receive a more inspired name in the not-so-distant future, and business.view, only available online.

Not every section has a column: The Americas, Middle East and Africa, International, Books and Arts, and Science and Technology make do without one. (These sections are typically a bit shorter than their counterparts with columns, except for the Britain section.) I believe that The Economist should have a column about science, engineering and technology. This would be the perfect way to start a conversation about science and engineering, and promote scientific literacy.

Technology plays an increasing role in today's world and affects everyone, sometimes in an indirect, unexpected manner - a fact emphasized in the fictitious case study of Harvard Business Review's October issue, where hackers find their way into a hospital database and ask for a ransom in exchange for patients' electronic medical records. People cannot question, say, the safeguards in place regarding their own data as patients if they do not know which questions to ask.

With respect to science and engineering, the issue of food-borne illnesses, for instance due to E.coli contamination, which is described in this recent article in the New York Times (you might never eat a hamburger again), or the research topics investigated by Technology Review's "35 Innovators under 35" would make great starting points for columns. People who are neither scientists nor engineers rarely discuss these matters, but benefit from research advances - and are affected by epidemy outbreaks or security breaches - all the same. The Economist has never shied away from presenting complex, technical topics in its finance section; it is time it took the same approach to science and engineering. I even thought of a name for the column - Da Vinci. Of course. Who else?

Second-hand (Stolen) Kindles

The September 6 issue of the New York Times has an interesting article on Amazon's efforts, or lack thereof, to track down the new users of Kindles reported stolen (Gadget Makers Can Find Thief, But Don't Ask). I wrote about the Kindle back in March; that post was about the prices Amazon charges to download new books.

The use of wireless technology should make it easy to locate lost smartphones, satellite radios, and e-readers such as the Kindle, which has the "ability to download books wirelessly and store hundreds of titles on a single device." Surprisingly, the NYT journalist writes that: "the gadget makers [...] often know exactly who has a missing or stolen device, because in many instances it has been registered to a new user" but have not acted on this information. "Even a request to simply shut down service — which would deter thieves by rendering their pilfered gadget useless — is typically refused." 

I was stunned by Amazon's reluctance to take action on this issue. The article mentions that its stance is hardly unique among American tech companies, and that foreign-based companies seem more willing to intervene. Most damaging to Amazon's case, I believe, is its negative reaction to the very valid points a customer raised after forgetting his Kindle on a plane:

1. "[H]e asked Amazon to put the serial number of his wayward device on a kind of do-not-register list that would render it inoperable," to no avail,
2. He believes Amazon should add "another step in the registration process for secondhand Kindles", in effect sending an email to the previous owner asking to confirm he has sold or given away his Kindle. 

Amazon has so far turned a deaf ear on these ideas, preferring to retreat behind the need for a search warrant. The customer above points out that this is not practical in large municipalities such as New York City, where he lives and where the overworked police has more important things to do than tracking down stolen gadgets.

This is a great opportunity for Sony's e-Reader, although it underscores how the issues that matter the most to customers, when it comes to their technology gadgets, often have nothing to do with technology itself.

Wolfram Alpha: The New Search Engine

The July/August 2009 issue of Technology Review has a long article, entitled Search Me, about a new kind of search engine pioneered by Stephen Wolfram, the founder of Mathematica - a well-known "technical and graphical software for mathematicians, scientists and engineers". The search engine, which launched in May, is called Wolfram Alpha and has received significant attention in the mainstream press, including The Economist (The Search is On, May 14, 2009), and in engineering blogs such as Mike Trick's (Careful with Wolfram Alpha, July 31, 2009).

The article in The Economist describes Wolfram Alpha as "a search engine that computes answers instead of looking them up". Wolfram himself describes his brain child not as a search engine but as a "computational knowledge engine", best suited for queries that lend themselves to mathematical computations, such as counting the calories in a recipe. It also takes advantage of Wolfram's excellent graphics software.

As an example, if you type IBM and Apple into Google, you will return this: over thirty-nine million of pages where the names IBM and Apple appear and have been indexed by Google's bots as they crawled the web. But if you type IBM and Apple into Wolfram Alpha, you obtain that: a detailed analysis of IBM's and Apple's stock returns, complete with latest trades, fundamentals, recent returns, relative price history, performance comparisons, correlation matrix, and even six-month, one-year or two-year projections using random walk models.      

Wolfram's search engine relies on "curated data", i.e., data that a team of experts has "amassed, scanned, processed" to use in computations; this is in sharp contrast with the traditional approach, which collects as much information as possible and spits it out in the same exact form in query results. Technology Review states that only over the past year or so have search engines begun to evolve, for instance recognizing addresses, phone numbers or restaurants through the use of tags.  

More recently, Microsoft launched a new search engine, Bing, which is supposed to "organi[ze] its results in terms of relevant groups rather than a series of links" (Economist.com, The Invention Machine, May 29, 2009). For instance, if you type Paris, you will get this - notice the headers "Paris Weather", "Paris Map" and "Images for Paris". But if you type Paris in Google, you will get that, which, dare I say, is just as well, with its maps and images at the top, and its "news results for Paris", "book  results for Paris" and "video results for Paris" lower in the page. According to Economist.com, however, Bing is believed to have some semantic-technology capabilities, although it is not clear whether it is currently using them.

Besides mathematics and engineering, computational knowledge engines could be most useful for "specialist fields, such as medicine or law, where the terminology is limited". The Economist.com journalist imagines a day where a chemist faced with expiring drug patents will find other uses for his equipment by typing a query into a semantic search engine. While this sounds a bit naive (if everyone can find business opportunities by asking a computer, the new opportunities will quickly disappear), the journalist also describes a fascinating demonstration of "deep content analysis", the next step in web searches ("far more impressive than Wolfram Alpha"), which will help machines "understand the complete and unambiguous meaning of human sentences".

High-Frequency Trading

The July 23 edition of the New York Times has an outstanding article on high-frequency trading. ("Stock traders find speed pays, in milliseconds.") High-frequency trading is a computer-based approach to trading, which uses sophisticated algorithms to spot - and take advantage of - price anomalies before anyone else has a chance to act.

The key is speed, which is measured in milli- or even microseconds. What matters is for one's orders to be processed before competitors', which has triggered a "technological arms race", to quote an executive at NYSE Euronext. This idea of arms race is echoed in an April post by Rick Bookstaber ("The arms race in high-frequency trading") and in a July post by Tracy Alloway ("The Cold War in high-frequency trading") on their respective blogs. Bookstaber is the author of A Demon of Our Own Design: Markets, Hedge Funds, and the Perils of Financial Innovation. Alloway is a contributor to the Alphaville blog for the Financial Times.

It turns out that high-frequency traders can, for a fee, receive early notification of other people's buy orders, thus getting a sense of the market's direction. This practice is known as "flash orders". Although we are only talking about an advantage of thirty milliseconds, such a short timespan is enough for the high-frequency traders to start buying, at lightning speed (and hence still at a low price), shares of the stocks others want to buy, and resell them at the higher price created by the high demand they just caught a glimpse of. The NYT article gives an example using Broadcom, a semiconductor company.

Senator Charles Schumer recently proposed banning flash orders in a letter to the Securities and Exchange Commission. ("High-frequency trading faces challenge from Schumer", Bloomberg, July 27, 2009) He "raises the stakes in a debate over whether computer-driven trading by hedge funds and Wall Street firms gives them an unfair advantage over other investors."

Goldman Sachs is noted in the Bloomberg article as not using flash-order systems. High-frequency trading, however, is said to account for a sizable part of Goldman's profits ("Rivals play catch up as Goldman thrives", Wall Street Journal, July 13, 2009); this fact came to the fore a few weeks ago when a former employee was arrested for allegedly stealing trade secrets before changing jobs. His expertise was in, you guessed it, high-frequency trading. Alloway of the Financial Times mentions the case in the post I linked to above, as well as here and here; more information can also be found here and here. The ex-employee "told the FBI he “only intended to collect ‘open source’ files on which he had worked, but later realized that he had obtained more files than he intended.”" (from this source)

This at least sheds some light on Goldman Sachs's extraordinary success despite the crisis. But the reliance on computer codes for competitive advantage and the recent accusations of theft, which GS says could cost it billions (because the software was uploaded to a server, where others may have found and downloaded it) also raise the question of how to protect key intellectual property. The days where traders only relied on hard-to-steal gut feelings are gone. 

Cyberattacks and the two Koreas

The New York Times published an interesting article last week about the "denial-of-service attack that briefly knocked offline the Web sites of some United States and South Korean government agencies and companies" during the Fourth of July weekend and the following week ("Internet's anonymity makes cyberattack hard to trace", July 16, 2009).

The attack has led to "South Korean accusations that the attack had been conducted by North Korean military or intelligence agents." What impressed me most in the article was the cautionary words of cyberwarfare specialists: "attributing the source of cyberattacks and other kinds of exploitation is difficult at best and sometimes impossible"; despite the rumors, "the consensus of most computer security specialists is that the attackers could be located anywhere in the world."

Experts emphasize that responsibility for this kind of attack is extraordinarily difficult to establish; when the culprit is caught (which rarely happens), it is usually because he made the mistake of talking about his actions - read for instance the paragraph about the "15-year-old Canadian high school student who was identified as a suspect only after publicly bragging about [high-profile denial-of-service attacks against Amazon.com, eBay and the like in 2000] in an online forum."

The difficulty in tracking down cyberattackers opens the door to potential abuse; government officials in fragile democracies (or plain authoritarian countries) could orchestrate an attack against their own websites, blame their neighbor or main enemy, and create a backlash against that other country to manipulate public opinion. One could also imagine a situation similar to the Gulf of Tonkin incident during the Vietnam War. Lyndon Johnson used the August 4, 1964 alleged attack against USS Maddox "to persuade Congress to authorize broad military action in Vietnam, but most historians have concluded in recent years that there was no second attack [the first attack was on August 2]" ("Vietnam Study, Casting Doubts, Remains Secret", New York Times, October 31, 2005) Experts say Johnson did not deliberately mislead the public; intelligence analysts had misinterpreted Vietnamese communications and, when they realized their mistake, covered their tracks by distorting the information instead of coming clean (source: NYT). About fifty-eight thousand American soldiers died in Vietnam.

One could also imagine a cyber-attack being incorrectly attributed to someone who has nothing to do with it. For instance, a Democrat running for reelection could accuse of his Republican challenger of bringing his website down during the electoral campaign, when it would in fact be the handiwork of his teenage daughter's disgruntled ex-boyfriend, who doesn't care one bit about politics but enjoys computer science. (I'm making that one up, but you get the idea.)

To give one more example, Estonia was the victim of a large-scale cyberattack two years ago and publicly blamed the Kremlin, which denied the charges. (The attacks happened after Estonia decided to move a Soviet World War II memorial away from downtown Tallinn, Estonia's capital.) To this day, Estonia has been unable to provide evidence to support its hunch, and one can assume it tried very, very hard. But the cyberattack could also have been the brainchild of computer whizzes who acted without the Kremlin's knowledge. After all, about a third of Tallinn's population is Russian.

It amazes me to think that even the top computer experts in some of the most technologically advanced countries - South Korea is known for its consumer electronics and Estonia has one of the highest Internet penetration rates in the world - have been unable to pinpoint the true source of their respective cyberattacks so far. This speaks volumes about the complexity of the World Wide Web.

Car windshields and Styrofoam cups

Diandra Leslie-Pelecky, guest blogger at Cocktail Party physics, has an interesting post, complete with equations and back-of-the-envelope estimates, on whether a Styrofoam cup can break a windshield. Her instinctive answer is: of course not. It's a Styrofoam cup! Then she saw the pictures - she was asked the question by a local reporter after someone in a car discarded a cup of soda through his or her window, cup that then smashed the windshield of a poor woman who happened to drive by in the opposite direction.

For those of you who can't get over the fact that all this damage was caused by a Styrofoam cup, the key isn't the cup (very light), but the soda that was in it (much heavier). As Leslie-Pelecky writes, "We know it was a cup of soda because, in addition to poor Marilyn [the innocent driver] getting wet, there is sticky dried cola-colored liquid all over the inside of the car.  The police pulled part of the cup out of the windshield (which is how we know it came from Sonic) and found a shredded straw inside the car." (The driver was lucky to be wearing glasses,which "were pitted from where pieces of windshield glass hit them.")

Leslie-Pelecky proceeds to estimate the force with which the cup full of soda hit the windshield, coming up with a range of 24-120 pounds because of the lack of data on the angle at which the cup hit the car and other parameters. It's a fun article to read for people who like science; Leslie-Pelecky points out that more kids might become interested in physics if high school exercises were more relevant to real life. But when it comes to throwing your cup of soda out of your moving car, the moral of the story is that it's a big no.

Stuff I'm Reading

• Guest column by Steven Strogatz about the Phillips machine on a New York Times blog - "It uses water to predict the economy." The first machine was built in 1949 and relies on dynamic feedback loops. (The article has a drawing of the machine showing the economic interpretation of the various inputs. "Phillips' Economic Computer, 1949", from the Science Museum, has a picture.) Fourteen Phillips machines were built, but only two are still working today. Although many dismiss the machine as a relic, Strogatz points out that "the Phillips machine foreshadowed one of the most central challenges in science today: the quest to decipher and control the complex, interconnected systems that pervade our lives." Also see "The computer model that once explained the British economy" (The Guardian, May 8, 2008).
• "Air France 447: A Detailed Meteorological Analysis" by Tim Vasquez is an amazing attempt to put together the weather conditions faced by the aircraft before it disappeared at sea. (James Fallows from The Atlantic mentions the study here.) While much of the technical language is beyond the understanding of non-specialists such as myself, the graphs provide fascinating examples of what scientists can do to aid investigations. Vasquez writes: "I did flight route forecasting for the Air Force in the 1990s. One of my assignments in summer 1994 was forecasting the sector between Mombasa, Kenya and Cairo, Egypt for C-5 and C-141 aircraft. The Sudan region had tropical MCS activity similar to this with little in the way of sensor data, so this incident holds some special interest for me as one of our C-5s could easily have followed a very similar fate. Using what's available to me I decided to do a little analysis and see if I could determine anything about the fate of AF447 and maybe through some circuitous, indirect means help give authorities some clues on where to look. " Let's hope that the study helps, ultimately, bring the families some closure.

A Convex Optimization Framework For Multi-Agent Motion Planning

Two weeks ago, I attended Jason Derenick's dissertation defense, entitled - you guessed it - "A Convex Optimization Framework For Multi-Agent Motion Planning". Jason was a graduate student in the Computer and Engineering Department at Lehigh, working under the supervision of John Spletzer; he has now moved to the University of Pennsylvania, where he is a postdoctoral fellow in the GRASP (General Robotics, Automation, Sensing and Perception) laboratory.

I served as a member of his dissertation committee, which is how I got to learn about his research. It turned out I had heard about him before - he was one of the team members behind the "Little Ben" car in the 2007 DARPA Urban Challenge. That car "was one of just six driverless cars to complete the 2007 DARPA Urban Challenge and it was the only car of the six finishers whose team had not received $1 million funding from DARPA to prepare for the race," according to Lehigh's press release.

In his dissertation, Jason has exploited recent advances in convex optimization to develop real-time algorithms for motion-planning problems, with an emphasis on optimal team formation changes and optimal collaborative target tracking. These problems lead to second-order cone programming and semidefinite programming formulations, respectively.

I was very impressed by Jason's work, and the wonderful presentations he makes of it. Of course, I only got to see the product of his research in its final stages, not the six years of hard work leading to it, but it is refreshing to bypass the difficulties and hurdles of graduate work for once (I've got my own students for that) and be shown the finished product without all the setbacks. While this is true of any committee I serve on, including for students in my own department, I particularly enjoyed the application to robotics - a departure of my usual line of work and a welcome reminder of my days as an undergraduate.

One reason why people - myself included - can get so enthusiastic about the work is that Jason excels at disseminating his results and explaining them at a high level anyone can understand (it is fair to say he is the National Science Foundation's dream research assistant). In addition to designing an excellent website, he has posted on his dissertation webpage an outline of his research and several videos and simulations of his algorithms in action. My favorite is the one with the aibos taking a delta formation, because everyone likes dogs, even robot ones.