Cars, Computers, and Interface Design
February 1, 2011 § Leave a comment
A few days ago, I was reading about how Apple could dominate car tech (if it wanted).
Two weeks of car shopping has driven home two big points for me: User interface matters. Car companies aren’t all that good at it.
Really? Because last I checked, car companies are pretty good at user interface. It takes me less than 5 minutes (in any car) to figure out to make it start, go forward, reverse, stop, turn left, right, turn on and off the radio, change the station, change the volume, turn on the air, heat, defrost, adjust the seat, mirrors, steering wheel, put on a seat belt, and start driving. And yes, there are a lot of buttons. So?
On the dream in code forums, one person asks a similar question:
A new car can have over hundred controls. For example, 10 controls for radio, 5-10 for heating and ventilation, 10 for windows, 10 for wipers, washers and so on. Most people can master these very quickly. Why is it that the car with so many controls is easier to use than a video recorder that has few controls. What makes the car’s interface so good and that of the video recorder so bad.
I love this answer:
…a car’s buttons are shallow in depth hence why there are so many, and a camcorder has fewer buttons that do more things (depth) in conjunction with other buttons.
The last thing I would want to do, while driving, is deal with some sort of menu tree when all I want to do it turn on my headlights.
Apple certainly does create some exceptional user interfaces – the iPhone is obviously a shining example of that. And yet 8 out of 50 states prohibit handheld use of a phone while driving, and 30 out of 50 states prohibit text messaging while driving. (http://www.ghsa.org/html/stateinfo/laws/cellphone_laws.html) This is because devices which require concentration to use necessarily divert your attention from the more pertinent task, which is driving.
Humans think and remember in physical space; we have a well developed sense of spatial memory. This is what allows us to type while not looking at the keyboard, pick something up off of the desk without looking, remember where we left our keys (well some of us), and so on. This is also what allows us to reach down to the volume knob and change the volume on the radio without significant distraction from the task at hand (driving). We know where the volume knob is, and we know which way to turn it. It is a physical object in physical space and our brains have evolved to be able to interact with such objects without using a significant amount of focus.
When you change that user interface so that there is simply one button to enter a virtual menu of options, sometimes many levels deep, we’ve suddenly made the task one that requires a significant level of concentration. We can’t rely on our built-in spatial memory any more, because multiple functions are mapped to the same physical space. At this point, we have to build a representation in our heads to keep track of where in the virtual space we are. Which app am I in, which submenu am I on, where are the buttons on the screen, and so on. We’ve gone from something as familiar and simple as a knob to something that requires (comparatively) a great deal of thought.
Of course, car manufacturers have tried to reinvent the automotive user interface before. In fact, several have them have even developed some very Applesque interfaces. BMW’s iDrive is probably one of the best known of these, and although it went through a number of revisions, and some users were quite fond of it, the overwhelming response was negative.
The design rationale of iDrive is to replace a confusing array of controls for the above systems with an all-in-one unit…iDrive has caused significant controversy among users, the automotive media, and critics. Many reviewers of BMWs in (automobile) magazines disapprove of the system. Criticisms of iDrive include its steep learning curve and its tendency to cause the driver to look away from the road too much.
This article on cnet discusses iDrive (as well as Audi’s MMI and Mercedes-Benz’s COMAND systems) in some detail. In the article, Wayne Cunningham rates each of the systems according to Jakob Nielsen’s Usability Heuristics. I find this interesting, as the second heuristic is “Match between system and the real world”, which is something that buttons and knobs are obviously quite good at, and nested menus of options are not.
Of course, if this type of design would function so poorly in an automobile, why do we love our phones, laptops, tablets, televisions, and video games so much in other facets of our lives? The obvious answer is that we aren’t driving while using them, and we want to be engaged with our technology. This should quite obviously lead not to the question of how to redesign the interface of the automobile, but how to redesign the process of driving. Since we are not able to use these devices while driving, we can choose to either design the device to suit the task (as discussed above), or modify the task to suit the device.
It isn’t Apple that is likely to succeed in this venture. The Distronic Plus cruise control system that Mercedes Benz is using is a step in the right direction. Google’s work on cars that drive themselves are even more so. Optimistically we have 10 years before cars that can drive themselves are available. At that point the game changes. The design of our cars won’t be constrained by the requirement that the driver maintain focus on the road.
Maybe then Apple will rock the world with its groundbreaking automotive interfaces, but until then, leave my knobs and buttons alone.