On the typography of flight deck documentation – NASA (1992) | Hacker News Reader

On the typography of flight deck documentation – NASA (1992)

ntrs.nasa.gov

211 pointsAudiophilip44 comments

Strange that I have worked with so many designers over so many years that ignore basic research like this. In ecommerce it is a constant battle against block caps for every title, every product name and every banner. It is also a constant battle to get copy in a legible font with sensible line heights.

Recently I switched some fonts from the designer fonts to native font stacks with sensible line heights and the conversion rate improved considerably, although I was not able to measure this as there were external factors, e.g. sales promotions, that make this not so clear cut. The consensus is that the faster, easier to read site is winning sales, nobody cares with such emotion about the exact font now it is totally different on all devices, things have gone from being exactly specified to whatever it happens to be on the native font stack, there is almost no way of knowing if that really is 'San Francisco' or not. What I have not explained to my critical colleagues is how the 'SF' font changes from 'text' spacing and font to 'display' spacing and font above 19pt. So the font automagically adjusts itself for optimal legibility and clarity. I think the NASA typography guys would have approved of this approach when designing for displays - use the best fonts on the device. I also think Tim Berners Lee would approve, the original web was supposed to be with the device doing the look and feel.

it is a constant battle against block caps for every title, every product name and every banner

I know how this feels. Designing for people with big goals and limited design intuition generally means battling the implicit, repeated guidance that EVERYTHING IS MORE IMPORTANT THAN EVERYTHING ELSE.

Frequently the decision makers are older people with presbyopia who are convinced that ALLCAPS is best when the better solution would be a larger font.

For the new air traffic control centre in Swanwick here in the UK, I worked with a team doing UI design for the big 20x20 displays that the controllers (ATCOs) would use. Because these were effectively radar screen replacements, the ATCOs insisted that the screens should be black, with all the aircraft tracks, track blocks and other such on the screen in bright colours, as this is what they were used to. The UI experts said no, research indicates that it is better (more restful on the eyes etc) to use a light background, with the tracks in dark colours, and pastels for the track blocks. It was interesting to see how much research went into all this, much more that simple user preference.

The upshot however was that the ATCOs won, and the radar-style colour scheme was what was delivered.

Cue discussion about terminals and all-nighters preferring dimmed colors.

Guilty of this. Black background and light grey, deep blue highlights, deep red syntax. Staring at white boxes too long makes my eyes hurt.

my firefox & terminal (& wallpapers) are all dark

I really cannot bare to read anymore in sites with bright background

actual dead-trees are a different beast :D

I would prefer the ATCO responsible for my flight not to be distracted by unfamiliar (or from his point of view outright hostile) color scheme, thank you very much.

I've seen some newer style (often grey/blue/wight on grey) tower / control radar screens and they seem so much better than the old black ones.

I have to ask why - if all that research was done, they settled instead on the familiar preference?

Are you not assuming that research is infallible and settles such a controversial matter once and for all?

The research in this case simply did not accommodate the preferences of the intended user.

I'm assuming that the research is intended to ensure fewer planes crash, so yes.

It is not clear from the post whether it was not just general UI experience.

How many planes crash in the U.K. due to ATC errors?

One is too many, especially when the fix is basically free.

You cannot be that certain, as you cannot rule out unanticipated downsides. The experience and judgement of the specific users in the specific use cases counts for something.

Well if the expert said that the research shows it, it must be so.

There is also this one: Legible, are you sure? An experimentation-based typographical design in Safety-critical context http://lii-enac.fr/articles/vinot-chi-2012.pdf

Wired had an article a few years ago: http://www.wired.co.uk/article/aircraft-typography

I've been writing a book on ways design can play a critical role in our lives. This reminds me of one of the stories in the book: A pilot cashes a plane because he didn't realize which decent mode he was in and 3.3 and 33 look awfully similar in the display. These things can be the difference between life and death.

For those interested in more: http://shop.oreilly.com/product/0636920038887.do

In schematic design, it is a common practice to use a letter in the place of the decimal point. For example, a resistor with value 2700 ohms is written as 2K7 ohms, or a capacitor wiht a value of 2.2 uF is written as 2u2F.

It's more than that. Medicines should be in bottles where the category is indicted by the bottle shape and color, and the stronger the color the more concentrated/dangerous the medicine is. This would reduce a lot of hospital accidental deaths.

Bottle shapes indicating content have a history going back to wine bottles.

In most small aircraft cockpits, the lever for flaps features a little rectangular "flap", while the lever for the wheels consists of a little wheel.

https://twitter.com/mmsBA/status/731040643674472448

https://www.quora.com/Why-is-the-landing-gear-lever-in-an-ai...

Edit to add: Apparently it's called Shape Coding, see the inventor https://en.wikipedia.org/wiki/Alphonse_Chapanis

For people interested in this, there are examples on Twitter

https://twitter.com/EzDrugID

https://twitter.com/ismp1/status/840239328467001345

https://twitter.com/TerryFairbanks/status/848743485213724672

https://twitter.com/EzDrugID/status/652036570313154560

The medical industry's resistance to this (and checklists) is a shameful travesty.

Absolutely. I loved Atul Gawande's book The Checklist Manifesto, about how to transfer safety lessons from aviation to the medical industry. Airplane accidents are somewhat more visible than individual deaths in hospitals, so the focus is clearly there, but I'm sure more people die of mistakes in hospitals than in airplanes.

https://en.wikipedia.org/wiki/The_Checklist_Manifesto

I always find it odd that Americans often don't write a zero before the decimal point, if a value is between 0 and 1. Like .1

It's easy to be confused with 1 and looks like a typo to my German eyes. I much prefer 0.1

Btw., we would write: 0,1 I think this is even more clear.

A side question, do you know the Germans always used a comma as the decimal point, or was it changed over because it was clearer.

The history isn't entirely clear. Early on, a lot of separators were used, but it seems that in the 18th century, France standardized on the comma, spreading it to the rest of continental Europe, while Britain settled on the decimal point.

And their spheres of influence pretty much settled the question internationally: https://upload.wikimedia.org/wikipedia/commons/a/a8/DecimalS...

It's somewhat frustrating how long O'Reilly books can spend in the early release phase.

I've been waiting for this one for a year now. It looks like an excellent read on the topic of bad design.

As one of the authors, I don't suppose you can shed some light on when it might be "finished"?

Thanks for the book. Just read the introduction and will be recommending it to my sister (worked once as a test pilot, will work in that field again). I've also started a lunch & learn thing in my office, I will probably be recommending excerpts from this as reading material in a couple months.

For anyone interested, it is also on Safari Books Online if you have access to that.

Related: the Typesets in the Future blog.[1]

[1] https://typesetinthefuture.com/

That blog looks great! Thank you for sharing :-)

For what it's worth, the US military often references MIL-STD-1472 [1] to frame general design requirements for pretty much anything that involves direct human interaction.

[1] http://quicksearch.dla.mil/qsDocDetails.aspx?ident_number=36...

This is really cool!

Just glanced through it, and saw this:

    5.4.2
    Orientation.
    Labels and information thereon (i.e., words and symbols) shall
    be oriented horizontally to read from left to right.  Vertical
    orientation, reading from top to bottom, may be used only when
    labels are not critical for personnel safety or performance and
    where space is limited. Where vertical orientation is necessary,
    the characters shall be readable in an upright orientation as
    depicted on figure 31. Sideways text and text read from bottom
    to top are unacceptable.

Such a definition seems like common sense, and then I realized there is probably someone out there in the vast military industry who probably doesn't care enough and may actually place a label sideways from bottom to top...

It is also amazing how much longer it takes me to read a label when letters are arranged vertically but not sideways. Sideways and vertical is quicker for me.

That is interesting. I agree with you on the speed of reading, but it does offer one very interesting advantage - every single label offers an indication as to the "proper" orientation of the object you're reading the label on.

I imagine that alone would be quite useful in emergency situations, or even in general loading/unloading operations.

My wife works for a shipping company, and unless it's on a pallet, it is likely to be orientated incorrectly on at least one of the legs of its journey.

I just had a quick scan through it but fascinating the depth and breadth of criteria. There's a table with about ~60 different room environments from lounge, galley, to actual workstations and the level that the room should be illuminated, and for workstation rooms, the reflectance of different materials in the room.

Great find!

ps I'm probably the only person here using safari, but the page to load the PDF didn't load for me so here's link to latest revision: http://quicksearch.dla.mil/Transient/72FC2386C1C2425E8298D6C... (2012).

That URL says "Transient" for a reason. ;)

There are a few other reports on the topic in the NASA archives. Such as this one which references the report linked to by this post.

Designing Flight Deck Procedures - 2005 - https://ntrs.nasa.gov/search.jsp?R=20110016464&hterms=Typogr...

This should have (1992) added to the title. Still interesting. What tipped me off was this observation:

There are over 2300 typefaces available today.

Note that this is from 1992. Even so, it's definitely fascinating to see the amount of attention to detail put into the design of the documentation.

As one (somewhat less serious) anecdote, remember the Best Movie mishap at the recent Academy Awards?

Better typography could almost certainly have prevented that:

https://medium.freecodecamp.com/why-typography-matters-espec...

The document is a bit equivocal on the use on serif vs sans-serif fonts but eventually states that sans-serifs are usually more legible. Interestingly, US pocket checklists (for use in flight) use a serif font for the routine checklists and sans-serif for the emergency procedures. I wonder if there was some research that supports that decision. ([1] is an F-14 checklist I found on the internet but other platforms' checklists have identical formatting.)

[1] http://www.avialogs.com/index.php/en/aircraft/usa/grumman/f-...

I wonder why they don't insert a small vertical space before each line with a bullet point in front.

For me, this makes it much easier to identify the groups of lines that belong together.

Reply on news.ycombinator.com