A magic box from GreenRoad Technologies (UK) will be installed in 200 military vehicles as part of a six-month trial to determine whether it's possible to detect bad driving and then intervene to improve it. As nearly as I can tell, the boxes have accelerometers (i.e., sensors that measure acceleration) that pick up how hard you accelerate, how hard you corner, how hard you brake, and how much hot coffee you spill all over your chinos. These measurements are used to give you continuous (red light, yellow light, green light) feedback on your driving. On the theory that if you give people a metric, they'll optimize it, the hope is that the boxes will improve soldiers' behavior on the road.

So What?
It's interesting to wonder what would happen if you just lied and installed a box that blinked randomly as the car was driven. It's the sort of experiment rat psychologists like, and the result is usually that the rat does whatever it's supposed to do even though the pellet (green light, in this case) is being provided without regard to its behavior. Maybe the researchers could save some money on accelerometers. (Come to think of it, in order for this to be a controlled experiment, they'd almost have to have a group of cars with random boxes. Hmm.)

For years I've been waiting for insurance companies to try this sort of thing, and I've heard vague rumors of research pilots along these lines, but never anything definite. Here's the idea as I see it: The insurer installs a box in your trunk and a needle on your dashboard. The box senses how safely you're driving and moves the needle accordingly. The position of the needle represents your insurance premium at that instant. Think of it as a speedometer, except it's a Big Brother-ometer.

I just ran across an article in the Wikipedia on prediction markets, of which I've long been a fan. Prediction markets' basic idea is to turn predictions ("Obama will win" and "McCain will win," for example) into stocks, then have participants buy and sell the two stocks on an exchange. The two prices will, if the market works its magic, reflect the relative share of the vote that'll accrue to each candidate. By watching the market, you should be able to learn the future. Simple. Ish.

So What?
This would be a very clever idea if it weren't so utterly insane. This was my reaction years ago when I first heard it, anyway. But it turns out that there's a fair amount of theoretical and concrete material to back up the claim. To your vast relief as well as mine, I'm going to skip the theoretical stuff in favor of the concrete.

The canonical example of a prediction market is the Iowa Electronic Markets (IEM), which for years (back even before the Web, if you can believe it) allowed you to trade on the outcome of presidential elections. Its record is remarkably good--it consistently gets the share-of-vote for each candidate correct when election night rolls around. (Of course, it's possible that it's just following the--usually equally accurate--polls rather than being driven by traders' independent assessments. There's no way to know.)

Accenture Technology Labs do (does?) two types of project: Accenture-oriented and client-oriented. Client-oriented projects tend to span a wide range of industries and are usually easy to write about. Accenture-oriented projects focus on systems delivery and often defeat my meager powers of description. This is pretty much one such. Read at your own risk.

Complex, expensive processes (assembly lines, refineries) tend to be heavily instrumented because the cost of a breakdown is high. So sensors are used to gather and analyze very detailed data on the process's evolving state. Large software projects are complex, expensive processes and the cost of "breakdowns" can be very high indeed. So where are the sensors? And where's the sophisticated mathematical analysis of the sensors' data? People fill out time sheets, sure, but that's about it. Where's the moment-to-moment visibility into the project's inner workings?

Haha! Twitter again (and again and again)! New Scientist is reporting on the use of Twitter by private individuals to coordinate and gather critical information during natural disasters.

So What?

Look, it's not my fault--Twitter just keeps coming up in interesting circumstances. If it were a publicly traded company, I'd probably buy stock (except that my holdings are deliberately limited to investment-grade chocolate figurines). Also, most of the stocks I bought jumped off a ledge shortly after I took delivery. So I wouldn't go totally by what I say.

Twitter is a micro-blogging service. "Tweet" messages of up to140 characters and anyone who subscribes to your feed can pick them up, both on the Web and on a cell phone. My colleague, a recently-married programmer and martial arts geek, has more than 40 people following his feed, only 10 of whom he actually knows. What are the other 30 getting out of watching the mutterings of a total stranger? Perhaps it's like listening to police radio--a little excitement to make up for what Tom Lehrer once called your "drab, wretched lives."

I recently offered advice on exploiting (parts of) Web 2.0's social aspects. The parts I didn't cover include crowd-sourcing and folksonomies, which I will cover today. (My qualifications in this area consist almost but not entirely of being a bemused observer of the 2.0 scene for the last couple of years.) Like the last piece, this is going to be packed. Hold tight.

Crowd-sourcing is the recruiting of large groups of strangers to perform work or create content, often for free. Wikipedia is crowd-sourced, as is YouTube. It's incredibly powerful and potentially lucrative--YouTube is now accepting advertisements, for example. Another story: The Threadless company crowd-sources T-shirt designs, crowd-sources the rating of those designs, then manufacturers the winning designs. (It compensates the designers). Threadless's founders made $20 million in their first year of crowd-sourcing. Crowd-sourcing is a very compelling proposition for businesses that want to create content--unfortunately, it can be very hard to motivate the masses to give you any content whatsoever, let alone the content you need.

A website called The Point ("Make Something Happen") provides a framework in which you can raise support for a cause. The cause may be as grand as getting an oil company to lower gas prices or as humble as persuading the local Starbucks to install recycling containers.

At The Point, there are two components to a cause: support and execution. Support specifies how many people (or how much money) you want to line up before action is taken (this is the "tipping point," in their parlance). Execution specifies the action you'll take if you get the support you need. So in the oil company example, the action happens to be "We will stop buying gas from this oil company," and the number of people required before action is taken is 20,000,000 (three have signed up). The Starbucks campaign is making more headway: the 60 people required will submit withering op-ed pieces to the local paper; seven have signed up to date.

Twitter...yet again. Stephen Rose over at Fast Company has discovered a new use for the beast: Following the activities of business contacts with whom you want to ingratiate yourself.

So What?

Twitter (skip this paragraph if you know) is a micro-blogging service. You use it by "tweeting" posts of up to 140 characters. You can tweet from your PC or your phone, and you can arrange to follow other people's tweets. If you're sufficiently interesting, others may follow yours. And that's all. Twitter is the most trivial important application you'll ever run across.

Display manufacturer NEC has released a report on monitor size and its relationship to productivity. Surprise: There's a positive correlation. Big surprise: for editing tasks (text and spreadsheet) the gains can be as much as fifty percent.

So What?

"Fifty percent"? Brighter screens, faster processors, ergonomic seating, better lighting, chair massages, "power" naps--all the things that are supposed to make us more productive--can't (apparently) hold a candle to the large display. And it makes intuitive sense: Having documents spread expansively around your screen is clearly much better than stacking your applications and hopping promiscuously from one to another.

There was a game played in the 1990s called Six Degrees of Kevin Bacon. The idea was to determine the smallest number of hops from any actor to Kevin Bacon, using movies as the links. So if Actor A was in a movie with Actor B, and Actor B was in a movie with Kevin Bacon, then Actor A's "Bacon Number" is two. (I regard the knowledge of pop culture needed to win this game to be nothing short of astounding.)

The game shows up in a more serious setting: If you need an introduction to a sales prospect (or anyone else from whom you want something), you have to have an intermediary--a person who can vouch for you and promise (if only implicitly) that you won't waste her time. Problem: How do you find an intermediary?

About a year ago I was told to develop a keen, spontaneous interest in the enterprise implications of Web 2.0, a contentious beast that continues to draw a lot of attention. This I duly did, and as a result I have lots of wisdom to impart. I'm going to focus here on the social (i.e., no Rich Internet Apps or mashups) manifestations of Web 2.0. Space is limited: Wikis, blogs and social networks are all we'll cover. Hold tight--this'll be terse.

Wikis are shared editing environments. The most famous is the Wikipedia, which weighs in at more than 2 million articles (compare the Britannica at 65,000). According to Don Tapscott, 2,500 people contributed to a Wikipedia article on the London bombings of July, 2005. This "brains per deliverable" ratio is outrageously compelling to business. Problem with enterprise wikis: Participation. Approach: Make it people's day job (don't ask them to contribute in their "spare time") and (to get their feet wet) have them create a wiki deliverable on a deadline rather than a "living document" with no due date. (The "spare time" approach can work in a sufficiently large enterprise--if you have thousands of people aimed at a wiki, some of them will probably contribute something.)

New Scientist is reporting on new gaze tracking technology designed for use in 3D virtual worlds. Gaze tracking has been used for years by people with motor neurone disease, cerebral palsy and other "locked-in" syndromes, but only to operate desktop interfaces. This more recent technology will bring the likes of Second Life and World of Warcraft to people who can't play them with a keyboard and mouse.

According to the article, the magic is a suite of "eye gestures" that includes (among others, presumably) glancing off-screen in pre-defined directions. The gestures let the eyes become more expressive than they are when they're used simply as a mouse. This has obvious advantages for people with severe motor deficits.

According to The Register, there were 1,400 incidents of crossing guard abuse (driving past while they're in the road, revving engines, shouting epithets, etc.) reported in the UK last year. Dozens of guards (they're called "lollipop ladies" because of the signs they carry and because, apparently, few of them are men) have been hit by cars.

Local councils are responding with 1) training and 2) the Routesafe Monitor, a double-headed video camera installed in the lollipop person's stop sign. The camera is activated when the sign is held up and monitors the situation before and behind. Anyone misbehaving while the sign is upright will be taped and (presumably) tracked down and remonstrated with.

The (narrow but very important) problem: Test scripts used for version 1.0 of an application will probably break when applied to version 2.0 of that application. Testers try to edit old test scripts so that they won't need to create new ones from scratch, but the process is slow, tedious and error-prone. It involves running a script until it breaks (this may take a while), figuring out whether it was a test script error and, if so, correcting it, and then starting over...potentially hundreds of times.

An example: Imagine what would happen to a test script if version 1.0 of the application used a combo box while version 2.0 used a text box for the same function. Obviously, any script that touches (what used to be) the combo box will break in 2.0. This is an example of a Changed Object (CO) error. The other type of error is Wrong Path (WP), which occurs when a script tries to access a Graphical User Interface (GUI) element (button, check box, etc.) that's been removed in 2.0. Either error will bring the test run to a screeching halt.