According to the current copy of Mercedes News (Winter 2003/4) DC and BMW are cooperating on developing internationally accepted standards for evaluating driver distraction. Quote:
"DC has been developing test procedures to assess a system's quality and ease of operation before it enters production. A joint project with BMW named ADAM (Advanced Driver Attention Metrics) has created a basis for international standards with which distraction potential can be evaluated.
"In the lane-change test, for example, the subject takes a simulated drive on a three-lane highway. From time to time road signs appear telling them (to change lanes while doing other tasks such as) to operate the radio, aopen a can..."
For all the criticisms of Mercedes and, to some extent of BMW, we have to remember that these companies do try to stay in the forefront of "driving" in general. I suppose everybody does something, but do they do as much as DC (and BMW)?
As one who participates in several international activities related to advanced intelligent vehicles and Intelligent Transportation Systems, I can say that BMW and DC are by no means alone in conducting research on this topic. There is a lot of motivation from the US government and the European Union, as well as within the corporate communities.
Most if not all of the car companies and Tier 1 suppliers are sending informed people to forums and committee working meetings, and are publishing papers from time to time. It's hard to say if any company does more or less than another, because such work is usually considered proprietary and hence may never be discussed publicly. An outsider might never see what's really going on.
I hope that Daimler/Chrysler and BMW conduct this research _legitimately_, make its findings readily available to the public, and act on the conclusions of that research in order to minimize driver distractions. If that happens, it should be the end to the iDrive.
For an interesting parallel in user interface (UI) design, do some research on synthesizers from the 1980s. When analog synthesizers were dominant, they had a single knob or slider for each function. This resulted in lots of knobs, buttons, and sliders. The UI people working for the synthesizer manufacturers had to be very effective in logically grouping related controls in ways that made them readily accessible, and which made interaction with them intuitive and "user friendly". Around 1983/84, the introduction and mass production of digital synthesizers changed the face of synthesizer UIs. Digital technology allowed a wealth of new features, including the ability to press a single button to call up one of over 100 preset sounds. The UI changed from a sea of knobs to an LCD screen, relatively few buttons used to select menu pages and specific parameters, and a single slider or knob for data entry (i.e., for changing the value of a specific parameter). This is a striking parallel to BMW's iDrive system, as it first appeared, where BMW replaced numerous knobs and dials with a single LCD screen, and a big aluminum knob that is used to navigate menu pages, select parameters, and enter values for those parameters.
Digital synthesis opened the door to a variety of features that were not practical or possible with analog synthesizers, but a curious phenomenon was observed over the following years: Musicians discovered that the new digital UI designs were extremely suboptimal for real-time interaction and manipulation of parameters. With an analog synth, if the musician decided to open up a low-pass filter during a performance, he simple turned a knob. In the new digital UI, he had to select the correct menu page (sometimes buried several menus deep), select the correct parameter, then change it's value. Digital synths became boxes that a musician programmed in their studio, and then called up presets during a performance while forgoing real-time manipulation of the sound while playing. Step over to BMW's iDrive as it was initially introduced, and we see that depending on the menu that is displayed, the driver may need to navigating through three menu pages to change a CD in his music system -- Quite a suboptimal interface for a person whose primary task should be maintaining vehicle control.
The synthesizer industry went through a second wave of UI philosophy by trying to simplify it's interfaces while still retaining the menu/parameter/data-entry approach. It did this by "simplifying" the menu systems to make commonly used functions more readily accessible, and by supplying additional *programmable* controllers. So now, while in the studio, the musician could spend some time to program a few optional knobs or sliders to do specific things under specific conditions (i.e., while I'm using this specific string pad, Slider-A will allow me to open up the low-pass filter, and Slider-B will increase the phasing). Still, under real-time performance conditions, the musician could not be inspired to do something other than what he had already programmed, and instantly be able to do it. Step over to iDrive 2.0, BMW's newest iteration as it appears in the BMW 5 series. iDrive 2.0, also affectionately known as "dummied-down iDrive", has reduced the knob's directional input from 8-ways to only 4, and it has its menu systems reduced in complexity with more commonly used functions made more readily accessible, all while still retaining the menu/parameter/data-entry approach. Even with this new version of iDrive, depending on which menu is displayed, you still have to go to other menus, select a parameter, and enter a value to change a CD rather than the old analog control that let you instantly press a "CD
2" button on your dash board to accomplish the same thing.
When you view BMW's mouse-like iDrive controller with this understanding of user interface history and design, and you consider that approximately 10% of humans are left-handed, you can appreciate how it makes a bad situation still worse: Imaging having to use your computer mouse with the "wrong" hand all the time. And in the United Kingdom, where drivers sit on the right side of the car, and where 90% of these drivers are right-handed and are forced to use an iDrive "mouse" with their left hand, you can add still more distracted and frustrated drivers into the mix.
The synthesizer industry finally smartened up, and they realized that they needed a completely different UI for synths in which real-time control was required. They ended up creating synths with front panels that had digital simulations of analog controls! The front panel was once again covered with knobs, sliders and buttons, logically grouped and organized, but now instead of having an analog knob that turned
300 degrees, you'd have an infinitely spinning "knob controller" in its place with an LED that "pointed" to the current position (parameter's value). It took the synthesizer industry years to realize that the menu/parameter/data-entry approach is virtually useless for spontaneous real-time entry of control information into a musical instrument.
Interestingly, Cadillac learned this lesson with their ill-fated Cimarron around the same time that the synthesizer manufacturers were entranced with their newfound digital gadgetry. The Cimarron had an "advanced" UI that let the driver control a lot of parameters from a single computer user-interface instead of a lot of analog controls. Customer demand for the product was weak, user acceptance was even weaker. Cadillac withdrew the product from production within a short time and chalked it up to "lessons learned."
Hopefully, BMW will be a quicker learner than the synthesizer manufacturers were back in the 1980s. Now if we could only get BMW to stop making ugly cars, we'd _really_ have something going :-)
-Steve Makohin | Reply to snipped-for-privacy@interlog.com '97 BMW 328 cabriolet | (hotmail acct is spam catcher) '98 BMW Z3 2.8 | '00 BMW R1100S |
When I first read about BMW's iDrive I was appalled, and greatly concerned that Merc was looking to go the same way. With lots of buttons at one's disposal (and I have a lot more in my new car than than in my old one) one sometimes has to think which one is the right one, but you can look at the road while you think. Once the right button is found you just hit it. Furthermore, most buttons don't need to be pressed in an emergency at higher speeds.
To take another practical example: my old, recently deceased telephone answering machine. Every function had its own button with a text label. This was replicated on the remote control, an accoustic coupler. Dead easy to use. A newer model I have -- never used remotely -- needs a series of beeps to activate certain functions. Terrible.
Mobile phones are another example, in my opinion, of devices that are becoming/have become over-complicated in an effort to provide ever more features. In my 'youth' calculators came out and we noticed all the 'extras' that were thrown in merely because it was technically possible and probably cost little extra to provide.
I agree with you that real-time control requires a completely different user interface. Thanks for a great writeup. Some of the things which MB has been doing are available online:
A general introduction at driver distraction studies recently appeared in Mercedes magazine. Note several linked articles etc.
formatting link
This Siemens Webzine "Looking into the Driver's Head" makes some correlations with accidents:
formatting link
Finally, a preview of the interface to come in future Mercedes, the 2005 S-Class, M-Class and R-Class. It will use a center mounted rotating wheel together with a center mounted LCD screen, to select and then press down on the wheel to select options. It is said to be more user friendly than I-drive 2.0, but I have not driven it yet. Credit to Sebastian Nast, who scanned the spy pics and produced this collage:
formatting link
And finally, one of my all-time favorite user friendly MB switches, the W126 seat switch which apeared in 1980, and which can be found in some books on software design today, as an exmaple to do it correctly.
Once I get stuck behind drivers on cruise control mode or in a traffic jam, everything becomes a distraction!
For me at least a GPS system that actively changes your programmed route and otherwise warns you of approaching traffic hotspots would be the best distraction-reducing device.
I then won't have to drive at nigh to enjoy a good drive.
I've been in Information Systems, application development, and design (and related fields) for over two decades, and one thing is an absolute certainty: Few companies employ the services of human interface specialists to help them define and design their man/machine interfaces. Most companies approach the MMI (Man-Machine Interface) only from a superficial standpoint in the context of "does it look good enough to get people to buy it." I'm confident they take the necessary precautions to prevent lawsuits against them.
Good user-interface design is challenging. It's time-consuming, and it's costly. The resources who are truly outstanding in the field are scarce, and it appears that the resources who are "pretty good" are either scarce as well, or other priorities override leveraging these resources to the greatest extent possible.
You can see by my sig that I currently own 3 BMWs. I'm looking for my next car, an attractive 4-place convertible that is larger and more luxurious than my current "loaded" 328ic. The only contender I see so far in the Mercedes CLK cabriolet (I have my eyes on the CLK55 AMG cab). The Bently cab is well beyond my reach. The two reasons I am looking at MB rather than BMW's new 6 series are:
(1) The MB is very attractive to my eye while I find the new BMW 6's exterior to be unattractive over-all, and boring from some angles. I also find the interior of the MB to be attractive while the interior of the BMW 6 is overly stylized, trendy, and inelegant. Views expressed are highly- subjective.
(2) The MB's user interface is much more conducive to real-time interaction. That's pretty easy to do when compared to the BMW 6: just scrap the iDrive.
I sincerely hope I get a CLK cabriolet in my driveway before Mercedes starts "learning from its mistakes" by going down the iDrive road. If that happens, I may just sit out the silliness in my current car and wait for premium car manufacturers to return to their senses.
-Steve Makohin | Reply to snipped-for-privacy@interlog.com '97 BMW 328 cabriolet | (hotmail acct is spam catcher) '98 BMW Z3 2.8 | '00 BMW R1100S |
I think the points made re: distractions are well-presented. IMO, the increasing intrusion of other things a driver can do in a car besides driving present philosophical major questions: -
What are all these non-driving-related activities doing to the act of driving? Are we trying to eventually make driving non-involving?
If manufacturers continue to stuff non-driving features into cars then what is the point to having drivers' cars that come with all this extraneous gear?
Should manufacturers enable and encourage more non-driving activity in passenger vehicles? Is that a safe thing to do? Likewise, should people eventually just treat private vehicles as customized rapid transit systems instead of an enjoyable and relaxing physical activity?
Do intelligent capable drivers really want more expensive, potentially unreliable features in their cars? Does the drive towards adding "features" reduce the offerings of "true" drivers' cars?
Do intelligent capable drivers really want machines to do "everything" for them?
Do we as humans "want" to depend more on machines or instead to retain and exercise more of our basic physical skills?
Given the lack of altruism in business, the naïveté of the public and the tendency for government safeguards to lag way behind applications of questionable technology, what is all this emphasis of "stuffing the car" turning us into?
How do we set reasonable limits on runaway pursuit of growth and profit?
Nice article. I think it presents the issues in a well-balanced manner.
I think a centre-mounted touch screen that can be hidden away would be better. A backup control using a rotating wheel would also be OK.
IMO touch screen controls would be much faster and more intuitive to use. I also feel screens should not be used by drivers while driving so they should be put away to avoid interfering with the driver.
I presume the keyboard is not part of the design being tested.
Credit
The visual seat-control mapping is not a bad idea but I think it does not work so well in practice. I think it only works well when the car is stationary. Many BMW drivers are more "enthusiastic" than MB drivers and are likely to re-calibrate their seat adjustments while in motion. I know the manual says you shouldn't do this but it's impractical to stop the car for seat adjustments and it's much less distracting if proper seat adjustments are made on-the-fly. For instance, when I am in the middle of tight traffic I prefer to sit straighter and closer to the steering wheel. Then when traffic clears and I have a long high speed cruise ahead I prefer to sit back, recline and extend the steering wheel. These situations often occur along a continuous stretch of road. You simply cannot stop anywhere and it is far safer to adjust your seat while in motion than to sit distracted by the discomfort for the entire duration.
The MB seat-mapped buttons are also in the wrong spot for ease of access when the door is closed. In a LHD, I believe you would have to use your right hand to cross over your chest to reach the buttons or, you would have to scrunch your left arm and turn your wrist at awkward angles to work the switches. Similarly, I think the seat adjustment switches in the E65 are of poor design.
I think the older switch layouts located low on the side of the seat as in the E39 and the MB W163 are better. They work by feel instead of by sight - you want to keep your eyes on the road. Your car tends to get pointed to where your eyes are pointed.
Yikes! An MB iDrive. I sure hope that it is used only to implement non-real-time control inputs, such as the programming of how long the headlights stay on after you turn the ignition off. Otherwise, well, I've said my bit already on this matter.
Yes, a classic design for an intuitive control.
-Steve Makohin | Reply to snipped-for-privacy@interlog.com | (hotmail acct is spam catcher)
When you look closely, you'll see that MB added buttons back into it. In fact they removed the shifter from the center console to the steering wheel (like in the 600 limo 1964-1980) to make more room for real-time buttons in the center. If the spy photo can be trusted, there will be four additional buttons located in front of the controller wheel and most likely these will be user *programmable* to demultiplex those real-time functions from the wheel/menus which the user subjectively deems important. Just as you described below for the synthesizers. I was also not pleased to learn that MB will use mouse/menu type interfaces. But it also seems the interface is starting to follow those steps you described so well for synthesizers, to digitally simulated analog buttons.
Some 15 years ago there was a type of entrepreneurship called "joint venture" in the late USSR meaning a company that had some western know-how and capital involved
Yup. In fact, even companies that make things like cell phones and car radios do extensive research like this. And they spend lots of money doing it right here at VA Tech:
formatting link
I have not heard of an international standard for driver distraction. However, this stuff is being researched so much that the patterns are probably becoming well known, so assigning values to them can't be far away.
The best one I've seen is the one in the Mini. I don't know if it's actually lacking anything compared to BMW's usual I-Drive -- it seems simpler, but also better designed.
Yeah :-) AKA "Increment/Decrement Volume Parameter Value" buttons ;-) There are books and white papers available on good user-interface design (non-machine specific, so it's not a "Macintosh" or a "PC" document). For people in the know, it is painfully obvious that there is a lack of understanding when certain user-interfaces appear. This applies to hardware as well as software.
-Steve Makohin | Reply to snipped-for-privacy@interlog.com | (hotmail acct is spam catcher)
This is probably why they're trying to develop a standard -- so they can point to it and say they met it, in case they get sued.
Yes it is. IMO this is the real challenge. The programming is just a technical exercise. No wonder it's all going to India.
Perhaps. Companies have not traditionally had these specific positions, so people rarely train for them. And since there's little demand, there are few training programs. They're just starting to appear now.
Yup. It's like the software documentation issue -- companies are too cheap to hire good technical writers, so everything gets left to the engineers, who are often lousy communicators.
Companies "get by" for years this way, but never know what hit them when they finally get clobbered by the competition.
Finally, we're coming to the end of the age where merely being able to produce these gadgets at all is innovation. The business has matured, and the ability to make stuff is itself a commodity. The innovation now is in outstanding design -- unless you can make things really cheap.
MotorsForum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.