Oh, very cool!

…a Burger King in Hillside, New Jersey tried out the MotionPower energy harvester from New Energy Technologies to see how it would hold up to the heavy traffic flow they experienced over the holiday period. As drivers wended their way to the window to get their Whoppers, the cars ran over a metal speed-bump affair that, using the weight of the vehicles to depress a plate and turn some gears, produced 2000 watts with each passing.

As it was just an initial durability test, the fast food franchise didn’t actually benefit from the “free” electrons but customers were treated to a very small light show that was installed to demonstrate the that system was working. The manufacturer envisions larger MotionPower machines installed in places where traffic is slowing down to prevent the scheme from requiring extra energy input from vehicles. [via autobloggreen]

Driving in urban parts of Israel is more difficult than in Los Angeles, in part because there is less rigid a distinction between roadway and sidewalk.

Having lanes that suddenly swoosh off in unexpected directions (while your direction becomes a “public transport only” lane) turns the whole thing into a kind of living labyrinth. The internal control tower dialogue goes something like this:

“So, if I want to get to Keren HaYesod, I can start out of Geula, cut through Davidka Square and swoosh around Agrippas. Just remember not to come out down Hillel, or there’s no right turn onto Keren HaYesod and I’ll have to go clear down to the Old City before I can start to come around again; I’ve got to sneak behind the Great Synagogue first. That’ll be fine. As long as I’m going to the bank on the right side of the street, not the dentist on the far side. For that, I’d need to begin my approach by coming through Rehavia.”

If you drive a taxi, of course, just go down Jaffa Road and turn right on King George V, which turns into Keren HaYesod. That would be too easy for the taxi drivers, though, so the municipality has thoughtfully dug up most of Jaffa Road for the past two years, just to even the score.

[picture from the Elms in the Yard blog]

Predicting user intention has a long history. There’s always the hope that you can train a computer to anticipate the user’s next move and launch the desired application or function at just the right moment, without requiring a user command.

The question is, how do you predict? How do you know what a user wants to do next?

The traditional methods can be generally categorized as:

1. Statistical methods. Study ten or a hundred or a thousand people using your program, and discover which functions are usually requested after which other functions. A common example: you might find that after launching Word, 90% of the time users next create a new blank document. Therefore, when launching the program, automatically cause a new file to be created immediately. Another common example: Apple Mail recognizes an email address format or internet link format in text, and automatically creates a clickable link within the mail body. There are mountains of ethnomethodological studies that try to provide relevant data for predictive use.
2. Track individual user habits. Allow the application to track a user’s actions, and learn the user’s behavior patterns. Then activate functions automatically based upon past use.
   A non-real-life example — more of a wish — from JK On The Run:

After I finish doing my email, or even before I’m done if there are too many emails to do them all, I want to go to Google Reader to check all the items from my RSS feeds overnight.  I can open up Firefox or just say “check the feeds” or the equivalent and the [Intuitive Interface] knows to fire up Firefox with the Google Reader page loaded.  The key to the learning capabilities of the [Intuitive Interface] is that just because I use Firefox doesn’t mean you do.  If it’s learned from your actions that you use Opera or Internet Explorer then that’s what it will use for you.  No overt training required, the [Intuitive Interface] can learn volumes about your preferences and what you normally do just by paying attention when you do them.  After just a short time of doing this the [Intuitive Interface] can be working WITH you, not just for you.  It will become a very intelligent personal assistant that works the way you do when you do.  It’s always watching what you do and WHEN you do it as most people’s work days are very routine when it comes to schedule.

3. Allow the user to control and register actions and preferences. Photoshop does this by recording your action history, and then letting you not only undo actions, but also “record” sets of actions for future application to other documents. The Mac OS does something similar in helping you set which applications are used to open which documents.

What everyone yearns for is something like the first two categories — where the user does nothing, and the computer comes up with the right action “like magic”. The problem is that in real life, only category three is really useful. Why?

Consider the following two reports:

One of the features on my three-year old Acura that I’ve come to enjoy is its keyless entry and ignition feature. Walk up to the car, touch a button on the door handle to unlock it, and start the car without inserting the key. All while the key stays in my pocket. It’s a feature now found on many cars and eliminates the need to find your keys in a pocket, briefcase or purse.

It can even tell the difference between my key or my wife’s. This can have some unintended consequences. If my wife enters the car first from the passenger side with her key, all of the radio stations and other settings default to hers. (She thinks that’s great as it reminds me to be a gentleman and open her door first.)

[from Phil Baker’s Concept to Consumer blog]

***

Blackberry has this nice feature where you type a word without bothering with capitalization or punctuation, for example, typing “im” for “I’m”, and it changes it on the fly. (Funny, because there’s no actual spell-check…) It’s a feature that’s convenient, although I tend to under-use it.

Anyway, little glitch, I tried to send someone my Israeli email address the other day. It ends with @netvision.net.il. Except that my alert Blackberry insisted it was @netvision.net.I’ll. I went back to erase/change/fix maybe 6 times, unsuccessfully. Not a helpful feature, in this case! Why should I be in a power struggle with my cell phone? […]

Found another one: can’t type the word “id” (as in Freudian), or the initials for identification or industrial design (ID). I just keep getting “I’d”.

When is the tradeoff of 95% accuracy offset by the 5% error rate (uncorrectable errors)? Another long tail question? Kind of.

[from Feature Power Struggle, posted in this blog]

You get the idea. I’m sure you can draw examples from your own life. Unless a use-case prediction is true 100% of the time, the frustration of an incorrect prediction has to be allowed for. If the error is minor or easily corrected for, then the predictive action may be worthwhile (eg, having applications create new documents at launch — closing the new document window is a minor inconvenience, and the extra wait is unnoticeable). If the error is harder to correct, or more annoying (How do you tell the car who is really driving? How do you override Blackberry’s auto-punctuation?), the frequent convenience may not outweigh the occasional frustration.

It’s worth pointing out that anything in categories 2 or 3 will benefit from unshared use of the device. Sharing machines/phones/computers/cars when preferences have been customized or learned for a particular individual will entail even greater frustration than if there had been no customization in the first place. Which leads us to more “Me”and less “We”.

[Disclosure: I work for Power2B, who are  developing a 3D touchscreen and interactive TV interface that predicts user activity by tracking actual trajectories in real time, rather than through any of the above systems.]

Part I

I’m from Los Angeles. In L.A., we say “You are what you drive.” Sad, but true.

In a sprawling city with inadequate public transportation and a high average income, people go everywhere in their cars.

When I was a child, the car was a vehicle. My grandparents had cars (one per family, not one per grandparent) with bench seats front and rear. I climbed in and sat between Grandma and Grandpa, and we talked while we drove. If it got hot, there were these neat triangular vent windows that popped open in back, and roll-down windows that took about five minutes to roll back up (and an equal amount of time to recover from). Maps went in the glove compartment. Eating and drinking in the car were not conceivable.

When I was a teenager, the car had evolved to become a second home. My parents’ cars (one per parent) each had air conditioning, radio with five station memory buttons, power windows, console storage and a cupholder. The map was a fat Thomas Guide stuffed in the seat pocket behind the front passenger seat. That was new, too: the front passengers each had their own adjustable bucket seat (slide forward/backward and recline). We didn’t talk as much together in the car; we listened together to the radio. Usually the driver chose the station, so it was either news radio (both parents), sports, pop psychology, country music, golden oldies (dad), or classical music (mom).

By the time I finished school, the car had become an extension of self, part of a person’s identity. When you got your first drivers’ license, you started thinking about getting your own car. When my dad handed down his (totally cool) car to me, the first thing I did was earn the money to install a cassette player. I didn’t talk to anybody (car phones were so new that I only knew three people who had them, all in their forties), except sometimes my little brother, when I took him to school. When my first children were born, we drove and sang along to tapes and CDs.

Today, car interior design has gone farther than ever in cocooning the individual rather than the group: DVD players front and back, separate headphone jacks, individual climate control (front and back; driver and passenger). Second- and third-row captain’s seats. Cup holders in every door, seatback, and floor panel. iPod and mobile phone jacks. Memory storage of your seat’s height, location, and degree of lumbar support.

“You are what you drive.”

Step back for a moment, and you’ll see that these design trends paint a larger cultural picture. The car has moved from a shared space (bench seats, no entertainment, little customization) to a highly personal space that is unlikely to be shared. When it is shared (for example, in a family mini-van), every effort is made to create as much private space as possible (individual seats, individual climate control, individual cupholders, individual entertainment and entertainment controls).

It’s a very different attitude about the car and how you spend time in it.

To me, it’s speaks of a very Western interpretation of mobility: Freedom, Entertainment, Movement, Privacy, Independence. Mobility = Individuality.

Part II

This attitude and the design ethic it inspired isn’t limited to cars. It has been the driving force (sorry) in mobile phone design for years.

Listen to the usability experts up until about a year ago. Everything was about how “personal” the mobile phone is. Studies showing that a high percentage of people don’t feel comfortable sharing their phone, or letting someone even use their phone briefly. The personal messages, notes, contacts, call history, browser history, photos.

Even more, there’s a sort of personal identification and relationship with the phone itself. Going further, your mobile phone number is more meaningful that your social security number — it’s one of your names. Your mobile number represents you, unlike a landline number which represents a location, and doesn’t follow you around.

I have a book here on mobile phones in Japanese life called Personal, Portable, Pedestrian. That pretty much sums up what UI thinkers saw as being important to users.

It’s all true, but there’s a big problem with all this: it’s all based on “Western” cultures. Cultures in which individuality, freedom and personal space are high on the list of life’s priorities.

Paul Adams (User Experience Researcher, Google) at the MEX conference pointed out in his presentation that if you look at countries like India and China, there are people everywhere.

Americans look at these teeming masses and say, “chaos”. But it’s not chaotic to non-Western eyes. What we perceive as “chaos” may be perceived locally as “shared space”. Paul gave the example of Southwest Airlines’ seating system, which for years was a “first come-first served” arrangement, proven to be faster than assigned seating. For Americans, this was perceived as chaotic. We prefer assigned seats because we place so high a value on our unique, personal, private space. It defines us. (Southwest has since changed over to a numbered boarding order — with it’s own adorable website to explain it. Which already tells you something.)

But that isn’t necessarily true of people in other countries, other cultures. What is valuable to one person may be undesirable to another.

Even in our “own” Western culture, feelings about personal space are changing. Definitions of privacy (personal secrets that you share with 800 blog readers…), of space (virtual, real, contained within a particular device or account) are changing.

Spaces that were once shared (eg., living room, public bus) are now personal (iPod as a “sphere of isolation”, killing time with mobile broadband). Spaces that were once personal (eg., Walkman music player, internet browser) are now shared (sharing headphones, Zune WiFi, Facebook Wall, IM, location based services).

Part III

I just got a Nokia 1208 as a gift; an upgrade to my Kosher Phone account. (We’ll talk about kosher phones another time. Suffice it to say for now that a kosher phone is a phone with no data capabilities.)

If you’re reading this blog, a dual-band Series 30 phone [bet you thought S30 was extinct in the wild] probably isn’t on your mobile tech wish list. It’s three main selling points are:

* Get instant access to phone features [one programmable softkey]
* Add a little color to your life [lo-res color display and exchangeable color faceplates]
* Monitor and manage your costs [calling card tracking and call timers]

Notable features on the 1208 are an integrated LED flashlight on top (where some phones have the power button or IR window), speakerphone, support for multiple user contact lists, pre-loaded polyphonic ringtones, dust-resistant keypad, durable materials construction with non-slip backing, and a very long battery life (7 hours talk time / 15 days standby).

You may have noticed that the feature set doesn’t exactly match your checklist of phone features. That’s because the 1208 was designed for… well… less-developed countries. Despite the fact that its being sold everywhere (which is pretty interesting), priorities in the design were cost reduction, durability (many users, dusty climates) and sharability (if you can only afford to have one phone per household — or even per neighborhood — then making sharing easier becomes very important).

Which is exactly why the 1208 was on my Wish List; it’s a great example of the new attention being paid to read people, real cultures, real usage needs in the design of products and services. It has taken a long time to recognize that overall, we have enough feature. It’s about delivering them in a meaningful way, and hearing what people truly need.

It’s also about respecting other people enough to accept their own mobile identity definitions and priorities (family, community, participation, responsibility, communication), without trying to impose our “better” systems on them. In doing so, we honor others, while creating new design and product possibilities that benefit everyone.

Lacking radar detectors Iranian drivers have come up with a more communal, cooperative system of detecting cops. As we would approach a speed trap or hidden checkpoint oncoming drivers would flash their lights and then make an odd twirling motion with their fingers as the cars neared. Judging by how anxiously or quickly they spun their fingers you got an idea of how close the fuzz was, an action I observed in every section of the country I visited. This ‘us against the police’ mindset offers a more revealing insight into how people feel about the authorities than a stack of newspapers or months of research.

[From the oddly-inspired Axis of Evil World Tour]

This video demonstrating a BMW design concept philosophy has won my heart. It expresses so much of my own User Experience philosophy, and through it creates a design of such beauty it almost hurts.

[via reaction beta]

Where should the GPS navigation screen go in a car? Drivers put them in all kinds of places, trying to find a spot where they can see the screen quickly and easily, without blocking too much of the window. Mounting the display on the dash (near the radio controls, for example) takes the eyes too far from the road when viewing. Mounting it on the windshield creates forward blind spots.

Where “should” the display be? As a projection superimposed on the view in the windshield, seems to me. Especially if it could be quick enough to superimpose the information right onto your view of the real world in realtime.

Other random thoughts about in-car GPS devices:

Here’s a view of the dashboard of a Tokyo taxi:

Is there never a point at which one says “enough”? It sure looks as though there is way too much going on within the driver’s field of view. Worse, look at the GPS control:

Yeah, that’s a remote control you see there. A remote control for the dash-mounted GPS. Now is there enough going on for you? (If there isn’t, try sympathizing with the little cartoon character in the seatback ad just visible in the bottom left-hand corner of the picture. But I digress.)

In contrast, here’s a super low-tech taxi dash in Kyoto:

The messages appear to be printed onto pieces of paper and stuck into place.

Here’s one just for you nervous people out there:

The other day, the driver of the taxi I was in was playing solitaire on his windshield-mounted GPS. He mostly played while idling at traffic lights, but I did notice that the occasional move input while driving was also registered. (For you touch-screen usability obsessives — like me — note that it was a resistive touch layer, and he was pressing the cards using the earpiece of his sunglasses.) This did not inspire confidence.

What was the manufacturer thinking?

Shibuya’s main intersection:

The pedestrian overcrossings form an elegant full four-way crosswalk right over the street. Here’s a stairway detail:

There are stairs with integrated ramps in Jerusalem, too (to help mothers pushing strollers). My friend calls them “suicide ramps”. This is such a long one that it does look suicidal. Yow.

Seen Los Angeles:

Pretty funny, huh? You have to look closely to be sure there isn’t actually a golf ball half-way through the rear windshield.