Thread: Discussion: Best practices for sustainable UI design

We talking skin design?

Yup.

there has been some other threads on similar ideas. the primary problem so far is that the android/ipod interface is it requires most of your captive attention. you really can't do much with the interface without looking at the device..

cars are completely different-- icons are larger, and easier to press because anything in your car can't (or shouldn't) take your captive attention..

I think what works best with these types of discussions are sample images...it doesn't need to be a mockup of exactly what your would prefer just something to illustrate the ideas your proposing. Even screenshots from other applications or platforms really help illustrate a point.

I typed up a few things a while back...

http://www.mp3car.com/vbulletin/soft...nt-end-ui.html

Good input. Thank you

My hope was not to assert my point; rather open a discussion and get consensus (if that is possible) on the future of the technology.

I was using the Droid as a reference. It can be argued that a familiar interface that is inherently intuitive has the net benefit of popularity.

Not to pick on any ones work as I know that these people put a lot of heart into their designs; I just think they are trying too hard to make it cool with detrimental affects. For example; The DX5 skin for RR is out. My first impression is "way too busy". On one screen I lost count of the words. At 80 mph I can't see those words anymore.

carPCs in general and front ends specifically have always been the domain of hobbyists. That is about to change and History will report that what will be common technology in the future was started with hobbyists. I know this; in 1977 I built an Altair from scratch.

As is always the case, there is a deluge of ideas and attempts before a standard is formed that inventors get behind. I propose that OM embodies it and we promote it. With the right marketing OM will displace most front ends in its technology alone.

My design will be a turn-key offering for a very specific audience and will include safety features, not just IVI. So I am going in this direction anyway. I like what Justin et al have built and I think we should get behind it to build a solid core of support.

So; who has the floor?

I see improved driver safety as a two part process...assuming the idea of a visible screen remains:
1) Alternative methods of interaction to minimize driver distraction
2) Proper UI design to minimize driver distraction

For pat 1)
Examples would include speech recognition/voice control, gestures, steering wheel buttons, etc. While OM has improved the first two alot to prevent a user from needing to take their eyes off the road, theres still room for improvement and certainly room for additional ideas.

Part 2)
This area is not so clear cut. When rolling out a UI design, we do driver distraction research, which involves everything from button count, to user focus timing to real world testing. Now obviously counting the number of milliseconds a user is distracted and the number, size and position of where he needs to press a touchscreen can give a lot of info into how safe an interface is. Unfortunately, every users usage patterns will vary and while we try to optimize as many as we can, aesthetics can sometimes be tough to balance with "the safest design". Based on our research you can see that new skins:

• do not use text except where absolutely necessary (universal icons used instead)
• the number of clicks to perform an action is significantly reduced (for any common task-no more then two)
• rarely used functionality is moved away from the users focus
• screen transition time, effect duration and especially load times have been significantly reduced

Research is ongoing but the best way to improve driver safety is to get as many ideas and as much feedback as possible and try out some new ideas.

I agree with most of the above, but I think a lot of it is overstated.

Just hop into a low end car and then into a high end one and you'll 300x more buttons and graphics and information. Does that mean that the guy on the more expensive car is in more danger of being distracted? I wonder how many more miliseconds it takes to change settings on the Acura, but still they did it anyway. I think it's perfectly usable, and I prefer it that way (more buttons with shortcuts, rather than menus and multi function buttons a la BMW - iDrive is the only reason they can still add tech and get that spartan look, which is overdone btw.)

For instance, Acura MDX:



Dodge Caliber dash:



I could go on and on. Lamborghini Reventon] clocks for instance, are far from simple, but they assume most owners that could use the info read the manual.

i have had an idea for a UI kicking around in my brain for quite some time.. and this sort of hits on the points i have been looking to attain. Problem is my extreme lack of programming and design chops. I'm a modder, not so much a creator, it's my lot in life.

But for me the perfect ui is an almost blank screen. In my imagination I see a central "window" that shows 1 icon per action.. you can swipe left and right to change that "action" or long press and drag to put a shortcut to that action on the main screen as a branch off of the central "navigation". The central navigation point would effectively be a portal into the action that has "focus". a doubletap would make the action full screen, and a doubletap would put it back in the "portal" or minimize it back to it's branch, still displaying data, but in a smaller format.

example:
you have music playing in the portal, but want nav, longpress and drag a branch over to the right and music keeps playing and is removed from the main portal. then swipe to navigation and you can see your route, swipe away and your route keeps tracking, BUT your music doesn't get interrupted by voice prompts, bring it back and it can stop the music to tell you to turn right.. etc. doubletap and it goes full screen and has the same interrupt capability. Then a phone call comes in. nav drops back into the portal and a branch for your phone is generated. right side of the "branch" answers the call, left side declines it, but it's just a picture of the person calling, or a phone if no picture. the call ends and the phone branch kills it's self and nav goes back to full screen.

i have a strong feeling that with OM this is all possible. It's just up to someone to create something like it.

voice commands could be used easily in this type of setup. make a full screen swype start listening, then say what branch you want. double tap the branch if you want full screen, or swype and say "end call" or whatever.

in my minds eye this is a globe with smaller globes being created off of the main with lines going back to the main globe.. but really a "branch" could be as simple as a button or whatever. the main globe could be recreated for multiple screens and branches could be "tossed" between screens to extend multizone functionality. say a backseat driver finds a burger joint they want to go to, they could "toss" a branch with nav to the top of their screen, which would indicate the "front" screen. if they wanted ot back, they tap the "line" that goes off the screen to drag it back.

I dunno.. pipe dream? wishful thinking.. or just too freakin much?