Monday, August 18, 2008

Design of dials and doors

Check out this control setup in a new Ford truck in which I was recently riding: Ford F150 Center Console Ford has, thankfully, gotten out of ovals everywhere design, and seems to have replaced them with circles everywhere. Note the five actual dials below the radio controls. The far left one is the drive system (various 2 and 4-wheel settings) the center are the HVAC controls, and the right is... what? Seriously, I wasn't sure at first. 12v is the only position labeled. So I try turning it. No movement. Pretty rapidly it becomes clear it's not a dial. Tug on it (carefully, in case I am wrong) and it's a hole, a power port. In the past, this was the cigarette lighter, but now it's a power port. The only hints it doesn't turn are a little depression to the right, and a lack of an indicator line pointing to the "12v" setting. But the overall style embodies the "rotary switch" meme. What am I supposed to think this item does? I know exactly how this happened, too. For pretty much my whole career I have had to deal with the same issues. I don't know what it's called in industrial design, and even within interactive, names vary. "Visual design" is the most common, I think. Visual designers are tasked to add a style to the product. Usually, one that reflects the overall corporate brand, but always their mantra is aesthetic and consistent. My problem is when they try to common, universally-understood, design language with another set, in the name of consistency. This is a great example. Aside from communicating "rotary switch" vs. "hinging cover," I would consider there already is a design phrase for "power port." It's a particular knobby style, descended from the auto cigarette lighter itself. Within interactive, the most common cases are trying – or needing – to replace standard controls, like scrollbars or form elements. Either a stylistic change is desired, or something like Flash is used to create a large portion of an interactive element, and the design is proposed with non-standard checkboxes, or scrollbars that don't work by click-to-position or dragging. I have actually sat behind the mirror and watched such items fail users. For any designer, of any interactive element, consider the value of your design contribution. Is consistency of visual look more important than instantly communicating interactivity, and meeting user expectations, through use of well-known elements?

3 comments:

John Bossert said...

Makes me think of a commercial airline cockpit. Eighteen kabillion switches, knobs, dials, etc. One would not make a pull knob look like a rotary dial in such a circumstance, because you want the dohickey to be instantly recognizable to the pilot for what it is. You don't want the pilot to have to think through a designer's motif and discern if the thing is really a hinged switch, a turney knob, or a pull out actuator of some kind.

Steven Hoober said...

Well, of course, human factors started really with the design of aircraft cockpits, and otherwise studying pilot interactions with their aircraft. It's still a key focus area for human factors practitioners (not me, I don't have an actual degree in this field) followed by things like control systems in nuke plants, chemical plants and the like.

The USAF pioneered a lot of this research, and still apply it. Similar principles of analyzing behavior and systems is behind the whole aircraft carrier operations thing (impossibly dangerous, yet improbably accident-free).

Interestingly, the Army seems to be doing this in a very halfassed manner, despite lots of new technology. I got invited to a (call in) forum for the DoD with the head of whatever the Army science directorate is, and asked about this. Seems they are purposefully forgetting about the existing research, and will eventually (once it's far too late) get around to setting standards for this for ground based devices.

Anyway, back to aircraft: They still fail. Control knobs are not a big issue, but do kill people. Two great examples are the Trident crash (yes, in the 70s) where the flap and slat controls were disconcertingly similar and adjacent, and the RPV accident I blogged about earlier where the control sets are conditional so the "camera zoom" rate becomes the "engine throttle" lever, and if not pre-set carefully, the aircraft crashes. As it did.

Another issue is audible alerts. Lots of crashes have been exacerbated by too many alarms, leading the aircrew to general flusterdness, and to focusing on the wrong issue (loudest alarm) instead of discovering the core issue. Another is overlapping alarms. Good example is the Helios crash in 2005 (Athens). The cabin pressure system switch was left in the wrong position. An alarm warned the crew when the plane started climbing, but its the same alarm tone used for "takeoff misconfiguration" (flaps not down, etc.). Aside from this confusing the crew, alarms on electronic planes sound without visible warnings; you don't also see a "press set wrong" light and have to look it up. I guess because there are too many possible conditions. Worrisome, still.

Presumably people are continuing to work on it.

Anonymous said...

In my country it is not fashionable to drive truck, people here like to drive tiny car to avoid stuck on traffic jam. And of course low power consumption