It's hard to define “user experience” but they say “you know it when you see it”. I say, “you know it when you feel it”. User experience is more than visual. Perhaps the best urban description it to say it is visceral.

Recently, I was reviewing ideas with a group do user interface designers. We were discussing software applications for smartphones.

While the design patterns being discussed were important, what was missing was attention to the person who would eventually hold a device in their hand. All of us in the meeting had a smartphone and yet, there was very little thought to “would I like to use'this thing we are designing?”

Precision vs non-precision gestures

Mobile devices support a number of different input. While shaking, bumping, and rotating the phone can be used, the majority of input comes from tapping and swiping. Tapping is typically a precision gesture – tapping a button, selecting from a list, typing on a virtual keyboard. The user must hit a specific target for a specific amount of time and not move while doing it. Contrast this with a swiping gesture where the user can typically start in any of a large part of the screen and only needs to slide the finger (or thumb) in the general desired direction. Thus, swiping is a non-precision gesture – scrolling a list, sliding between two screens, opening a menu at the side or bottom, exposing a drop-down from the top (for examplr the iPhone notifications page).

When given the choice of implementing a precision vs non-precision gesture, it is “kinder to the user” to go with the non-precision choice.

Left vs right handed users

A smartphone is not a large device – even the comparatively huge 5″ devices. Most smartphones are designed to be usable with one hand (as seen in the photo). The user is cradling the phone and only has their thumb for interaction. The thumb naturally follows an arch from the upper corner closest to the hand to the lower corner farthest from the hand. The challenge is that these locations are dependent on which hand hold the phone. In the right hand, the ease of access starts with the upper right, then lower left, then upper left, and finishing with the lower right as the most difficult location to reach. A user holding a phone in the left hand would reverse these to upper left, lower right, upper right, and then lower left.

So where do you put input control when you have both left and right handed users?

Part of the answer flows back to precision vs non-precision input. The more non-precision input you achieve, the less challenging to left and right handed users. Next, the comfort of top buttons exceeds that of bottom buttons.

Conclusion

There is no perfect answer. However, by thinking about “how” a user will interact with a mobile application, (in addition to why and what-for) you will achieve a better user experience and a happier user.

This is a follow up to the previous post about already making an upgrade to the panel, even though it hasn’t actually flown yet.

Not including some mental perturbations, the upgrade from the new panel to the newer panel took less than a day!

The wire harness only needed one wire cut and that was not because of the new panel. It turns out I missed a ground wire in the autopilot servo harnesses and had to add those. Other than that, the wire harnesses were rearranged but nothing had to be reworked.

The new panel is probably not what most would have designed if they started with a blank slate. I would completely agree. It is a bit odd to see neither of the two largest screens in the center.

Esthetically, the finished panel is a little jarring but ergonomically, it actually works. Since the RV-8 has a stick rather than a yoke, the flight hand is in the center. That leaves the other hand to punch buttons. With the left hand on the stick, the right hand has the map right in front of it. Similarly, with the right hand on the stick, the left hand has the EFIS / autopilot right in front of it. While I tend to flight with might right hand, switching is not uncomfortable.

This panel change came about sooner than expected. I knew sometime in 2013 I’d likely replace the autopilot. I just assumed it would be late in the year. Well, TruTrak made a really good offer to trade in my obsolete unit. Along the way I also bought an old but unused standard servo and swapped out my big high torque servo. That switch netted me some extra cash which helped bring the new TruTrak EFIS SG AP3 into my price range. The challenge was going to be how much more panel space the new unit required.

I had to cut an addition 5/8" from the left side avionics space. I also needed to fabricate two new "shorty" rails for the remaining gear, along with a few small fastening plates (these hold the remaining its in the radio stack together for stability.

The radio moved up to the top and the old mechanical airspeed indicator was retired.

When I tally it all up I spent about 3 hours thinking about the project, 3 hours doing the metal work, and another 2 hours moving the wire harnesses around and reinstalling everything.

One of the design goals of the new RV-8 aircraft panel was to support upgrades without starting over. Even thought the panel is going to be new, much of the avionics are used. Some of those will need to be replaced over the next five years. The DFC-250 autopilot controller is old and costly to support so it will eventually be replaced. The FAA will mandate new transponders one or before 2020 so the old KT76A will need to be replaced. There is even the likelihood that the radio or GPS will need to be upgraded in the next several years. Even the ELT (emergency locator transmitter) will soon need to be replaced to meet FAA requirements (not yet in effect).

So it came as a bit of a shock that the first upgrade would happen before the panel ever flew.

It turns out the old DFC-250 was closer to its “end of life” that I knew. The manufacturer -TruTrak –  announced it was discontinuing support for the unit because the parts were not longer available – it was now impossible to repair. TruTrak was very accommodating with my situation and after a couple of phone calls and discussing my current usage, they made a very attractive offer to trade in the old DFC-250AS for a new EFIS SG APIII. The change also meant swapping a servo but with a little luck, that change will actually reduce the total cost of the upgrade.

So the panel will go from the current rendering to the new one (next two pictures respectively). The layout is not optimal but it will require the least amount of work to achieve.

I chose, early on, to label the panel using water slide transfer decals. My plan was to cut the panel and do all of the assembly work. Then disassemble the panel and finish it, followed by final assembly. For the most part (and without sharing excess gory details) that is pretty much how it worked out.

The finish process consisted of cleaning the aluminum with lacquer thinner to remove any residue, marking, oils, etc. Then a coat of etching primer followed by the base coat of light grey to match the old panel. I then printed out the labels onto waterslide transfer using my laser printer. After cutting out the individual areas, I soaked the decals. This is where I encountered the first problem. The laser toner washed right off when I tried to handle the decals. I soon found out why some tutorials describe an extra step – lightly spray the printed water transfer sheet with clear coat after printing and before cutting. This locks in the printing. Fortunately, I had planned ahead a bit. Since it is not possible to run a sheet of water transfer through a printer twice, and knowing I was likely going to make a mistake, I filled the 8.5×11 sheet with multiple copies of my various labels.

Once I had applied all of the labels, I let everything dry for a day. I then returned to the paint booth and shot three coats of clear coat, let it harden, and then did my first pass of black sanding with 800 grit. I was careful to only take down the high spots where the clear coat was on top of the labels. I then shot another three coats. This time, when I sanding with 800 grit, I was trying to level the paint surface – leaving a little clear over the labels and most of the clear everywhere else. Then I switched to 1500 grit before buffing with 2000 and 3000.

The final result is not perfect but you need to be within about eight inches of the panel and at just the right angle to see the edges of the decals.

I will admit a dry transfer system would have been even better but those systems cost $100 or more vs the $10 for water transfer.

Here is another trick. It turns out that Microsoft Powerpoint prints “inch accurate” output. I used that to my advantage.

I created templates of the switches and circuit breakers. (You can download my file here.) I did an accurate layout. I printed this layout and taped it to the panel. You will notice each object has an “X” on it. That shows the center. I then simply used a center punch on each “X” to set where to drill for each hole.

When it came time to print the water slide decals, I deleted everything except the text. This guaranteed that the labels matched the holes that were drilled.

Confession time … The above process does work. The mistake I made was placing label text too close to holes. In the template I created, I did not account for the size of the washers that on the front of each breaker and switch when it is installed in the panel. In my finished panel, the nice shiny washers touch the bottom of the text.

I was recently asked to describe the role of a user experience (UX) designer and how it was different from a user interface (UI) designer. I pretty much failed at the task. The best I could do was a sound bite …

UX is how a person feels; UI is what a person uses.

An article in Mashable on UX Mistakes provides some examples that help give meaning to my “ten work answer” …

People expect technology to follow the same rules as human-to-human interaction. Think about it, how do you decide whether someone is truly kind, or if he or she cares about you and is acting in your best interests? It’s not the clothes they wear or whether they have the right accent. It’s the little things they do for you, and whether they come through for you in a pinch. The same is true of websites.

The article is a lot more than just that one quote. Anyone who works in software design – UX or UI – will find a nugget or two worth remembering.