The wrong projection

On a recent flight from Amsterdam to Houston, I turned on the “moving-map system” in the in-seat entertainment system and was surprised to see that though we were halfway through the 10-hour flight, the map made it look like we were minutes from landing in Texas. Sure, I had dozed the delightful doze of the jet lagged international flight, but had I actually passed out? Or had time slipped by that quickly? Shouldn’t we be somewhere over Greenland? That looks about halfway.

Then I realized that it was the map itself that was to blame. The arc was being true to the plane’s path across the map, but with a map as distorted as this one, it was bound to be confusing.

Like most cool things, this gets nerdy quick. See, when you try and take something that’s pretty much a sphere (the Earth) and fit it to a rectangle (the in-flight entertainment screen) you’re going to run into some deformation. There are many, many ways to crack this mathematical nut (the awesome site Radical Cartography lists 30) and each optimizes some things at the costs of others.

The designer of this system had chosen to use the familiar Plate Carrée projection of the Earth. It’s ancient, and quite familiar to travelers. It’s used everywhere. So certainly, it optimizes for initial use. At a simple glance, the traveler knows what he’s looking at.

But this projection, in forcing the longitude and latitude lines into tidy squares, severely squashes areas that are closer to the poles. The result is that—unless you’re traveling from one point on the equator to another—an actual straight line across the surface of the planet will appear on this map as oddly arced. What’s worse is that the arc won’t mean the same thing across its length. Nearer the poles it will be stretched and closer to the equator it will be squished, resulting in the weird, jarring experience of watching the plane zip to Ontario, and then crawl to the Gulf Coast.



For a passenger on a plane this might not matter that much. But in more mission-critical operations (military, air flight control) this could mean costly mistakes in estimating time and space. As a designer, I couldn’t help but try solving it. What display would be less jarring? Thinking about my colleague Stefan Klocek’s Hierarchy of Effort, we can start with simple surface changes.

Mark time

The simplest visual improvement would be to put tick marks along the arc that indicate hours. They would still be squished near Amsterdam, but it would communicate the time/distance discrepancy of the current projection. The system wouldn’t need a lot of fundamental restructuring, and it would provide a cue to the traveler about the weird inconsistency.

Change the projection

Another, deeper, solution would be to use a different projection. As mentioned, there are lots of projections available. One that optimizes for keeping these “great circles” as straight lines on a map is called the “Space-oblique Mercator projection,” or SOM. Created in 1976, it’s a relatively new projection that uses some tricky math to ensure that the distance and the time remain consistent for the line between any two points. It looks unfamiliar at first, but provides a consistent time/space relationship, while maintaining the context of the whole world.

Unfortunately, I couldn’t find an online way to make a custom SOM map from Amsterdam to Houston, so I couldn’t comp it up for you, but here’s an image showing the great circle along the Prime Meridian projected as an SOM.

https://www.e-education.psu.edu/natureofgeoinfo/book/export/html/1541

SOM projections are certainly unfamiliar to most travelers, but it’s perfect if you want to keep the mental model of time and space being constant between departure and arrival destinations, all while showing a complete picture of the earth. There’s a slight problem if the plane is not traveling in a straight line, or deviates from a straight line, but minor deviations suffer negligible distortions.

The main risk to using a map like this is blowing people’s minds with the unfamiliarity of it. “What is this map?” they might think, “It looks like Earth, only it’s not.” (Only it is.)

Remove the deformation

If more fundamental changes were possible, showing a 3D globe would mean you wouldn’t even have the deformation problem at all. You could show it from a floating-overhead-camera point of view or a zenith-looking-down point of view.

Comps made from Google Earth with Adobe Illustrator overlays.

This floating-overhead-camera perspective uses a deeply familiar distortion, i.e. single-point perspective, similar to what people experience on a road or a sidewalk. The distortion isn’t removed, but is aligned so as to utilize the distortion of perspective. Much of the traveler’s prior experience directly maps to this view, and really only shows what’s of immediate interest: What am I seeing out the windows, and how much more time is left in the flight?

Comps made from Google Earth with Adobe Illustrator overlays.

The zenith-looking-down perspective shows both the departure city and the arrival city while acting as a progress indicator with a familiar kind of distortion, a simple sphere. 3D views calculated in real time require more powerful processing and more fundamental restructuring of the system, but many commercial moving map systems have real-time 3D rendering (Airshow 4000, JetMap, iXPLOR, Niceview, see below) already.

Or all of them

My experience tells me it’s time to abandon the Plate Carrée projection, but there’s a cost of familiarity. Many of these in-flight map systems cycle between different views, so new views could be included along with the more familiar ones, cycling in a loop that keeps the familiar map with its weird distortions while including these more informative and direct views as well. You could even include the SOM and label the various projections to encourage people to learn more about them, if they wanted.

A lot of the work of interaction design is about shrinking feedback loops, removing layers of interpretation that are unnecessary or distracting, letting more fundamental or reliable mental systems do the work. This is one example that at least for us travelers, stares us right in the face. Positive first impression through familiarity should not be the sole measurement of success, particularly on a system where the user has several hours to become familiar with it. Understanding a thing across the full spectrum of its use is often more important, and more critical.


Related Reading

1 Comment

Vanessa
It's about time someone commented on the in-flight mapping system. It's one of the few things that gives me comfort as I have a fear of flying, and I believe that if the map could more accurately depict where the plane was, and what were the surroundings, it would make me feel more aware of and in control of the experience. I hope you suggest this idea to some major airlines so they can consider adding a few extra map visualization screens to their existing system :)

Post a comment

We’re trying to advance the conversation, and we trust that you will, too. We’d rather not moderate, but we will remove any comments that are blatantly inflammatory or inappropriate. Let it fly, but keep it clean. Thanks.

Post this comment