THE IMPERFECT SPEEDOMETER 

(Click HERE to download the Power Point presentation of this paper) 

See also: The Instruments I don't like in my Car

(1) Warning- this paper was presented to the 3rd ‘The Eye and the Auto’ World Congress in Rochester, MI, 23-25 June 2005. I am indebted to the organizer of this wonderful congress, Dr Philip C Hessburg MD, for the opportunity he gave me to discuss the problems of the ‘unreadable speedometers’ with several other professionals of this sector. Ophthalmologists and ergonomic specialists agreed with my views. (Instrument panel manufacturers and car designers were less pleased.) This paper is designed to be read in conjunction with the similarly named *.ppt presentation that you can download from this same page. 

Abstract: the speedometer is the single most important instrument of a motor vehicle. The driver needs it to check the speed of the car and respect legal limits; being enabled to do this quickly and easily reduces glance time and mental workload. The classic analogue instrument with black dial and white needle provides an optimized layout with an excellent contrast and unsurpassed overall readability. It is ergonomically sound, cheap to build and easy to illuminate by night.

Dials with a light-colored background and dark targets (needle/numerals) may not be as ergonomically sound because their illumination at night will require a radical transformation of their appearance with a complete inversion of the background/targets relationship from day- to night- mode. Reading this speedometer is difficult during the periods of transition (dusk, dawn), when the black-on-white layout must change its configuration to a white-on-black one. During this transition, the luminance contrast between the various parts of the instrument falls sharply: as a consequence, the readability of the instrument will decrease and more glances of a longer duration will be required to read the data. When the luminance contrast falls under the level of the ‘minimally distinct border’ (MDB) the unit will become virtually unreadable.

During these periods of transition, the insufficient luminance contrast between the components of this type of instrument makes it difficult to read, thus increasing the driver’s mental workload. It appears therefore to be a risky choice, especially when considering the number of drivers (senior citizens, patients with uncorrected ametropia, lens opacification or corneal scars, refractive surgery patients, etc) that might already have a low contrast sensitivity. Until this layout is modified to optimize their luminance contrast in every condition of illuminance, canceling the low-contrast ‘transition’ periods, this choice does not seems an ideal one. 

DISCUSSION 

As an ophthalmologist, I have always studied the visual interfaces of the automobile, dedicating a special care to the design of the speedometer. (2) This instrument is, according to EC Commission Directives, ‘that part of the speedometer equipment which indicates to the driver the speed of his vehicle at any given moment.’

The importance of this instrument is clear. Among the many instruments that decorate today’s dashboards, the only two that are really important are the speedometer and the fuel level gauge; and the latter could be easily replaced by two or three colored warning lights.

‘Normal’ drivers support this concept. A survey conducted by Ethington, Steelman and Clark on 483 respondents showed that the data that drivers want to see permanently displayed are, in a decreasing order of preference, speed (96%), fuel level (88%) and time (84%). Oil pressure, rpm and other data, are perceived as being less necessary. The famed automobile writer L.J.K. Setright remarked that “….Miles per hour, on the speedometer, and the number of gallons of fuel left in the tank…are about the only numbers worth knowing while driving.” (3) The driver’s visual system must concentrate on the road ahead, (4) but the ever-increasing enforcement of speed limits everywhere means that the driver has to check quite often his speedometer: reading it quickly and accurately is important both for safe driving and for legal reasons. A permanently-readable speedometer is clearly required by the law. (5) This is an excerpt of the still void EC type-approval legislation, that clearly states that the speedometer display shall be situated in the driver's direct field of vision and shall be clearly legible both by day and by night.

The factors affecting the written communication process of this type of interface are comprehensibility, legibility and readability. These parameters are fundamental for the design of a good visual interface, and legibility is paramount in this case. We know that it is advisable to use white markings, pointers and numbers on a black background for displays to be used in reduced ambient illumination. We know the basic characteristics of the data that a speedometer must convey to the driver, its possible placement in the dashboard, the variable conditions of illuminance possible in the interior of a car and the needs of the visual system. (6) We also know that the driver must time-share the primary visual task and the secondary tasks. (7) As a consequence, the ‘proper speedometer’ is in fact a ‘classic’ analogue instrument with a diameter not smaller than 5”, (8) a black dial as a background, white or orange numerals and a full-length needle rotating from 8 o’clock clockWhise to indicate an increase in speed. Ideally, this instrument must be placed along the line of sight of the driver towards the road and high enough in the dashboard so that it does not require the eyes to turn down more than 15-20°.

This layout fits perfectly into our most common brain  ‘template’ for this type of instrument, as it follows stereotypes as old as instruments themselves. It offers an excellent luminance contrast at every moment of the day, it is easy to illuminate at night with reflected neutral, amber, green or red/orange light: the dial remains black, the numbers glow in any color that might be desired. (9) Even older drivers, who have usually a lower contrast sensitivity and would not read easily a lower-contrast instrument, can read this one quite easily. Having the same articulation between dark and light areas in every moment of the day and in any level of illuminance gives it a very good readability. (10)

(11) Moreover, retinal fields are known to be extremely sensitive to the smaller angular movement of a target shaped like a needle on a round dial; this improves the legibility of this type of instrument. Another functional effect of the peculiar orientation of the retinal fields is that the visual system reacts more quickly when a target is horizontal or vertical. A clever engineer can therefore design this dial so that the indications for the most important speed limits (in Italy, 50, 90 and 130 kph) will be found at 0°, 90°, 180° for maximum legibility. An indication placed on these axis will be easier to ‘understand’.

Sadly, ergonomic efficiency does not seem to be the only parameter by which a new instrument panel is judged today. An author has written that “…new applications in instrument panel designs…..will give automotive designers more opportunities to make their vehicles distinctive from those of other competitors.” Marketing has its needs, but we must also admit that ergonomic results of this kind of innovation are not always positive. We have tested hundreds of cars in the last 25 years, very often finding ergonomic errors of varying importance in their visual interfaces. Those errors were nearly always caused by the desire to offer new, ‘innovative’ visual attractions to the potential customers. 

 (12) A typical example of an innovation introduced for ‘style’ that can cause some potentially serious readability problem is the modern instrument with a white/silver/yellow dial. This is not a real innovation: cars used white instruments for decades, until a better design was developed. It has been recently re-proposed by several manufacturers whose interior designers look at white dials as being ‘sporty’ and ‘modern’, or, in other vehicles, as ‘classy’. (13)

When the dashboard lights of this instrument are off (daylight mode), the numerals and needle are dark. If we look at this instrument panel in broad daylight, contrast is as good as in the classic black dial/white targets layout: the dark/light articulation is simply inverted and the contrast seems fairly good, especially when the luminance of the dial is at its maximum and the needle and numerals look very dark (14).

Illumination is necessary anyhow to read the instruments by night, and it would be ergonomically unsound to maintain the contrast characteristics of these silver/light grey/white dials unchanged. In the darkness of the night, this would create an ample area of brightly lit dials in the lower/central part of the usable field of view and all this light could disrupt or anyway disturb the attaining and maintaining of the delicate condition of low-level photopic or mesopic adaptation that is so necessary for the primary task, that is driving. To make these light dials readable at night forces therefore the engineers to change their luminance characteristics. (15) As it is desirable to keep the dials’ face as dark as possible in conditions of low ambient illuminance, this means that the numerals and needles must be illuminated as light objects against a dark background. In today’s ‘backlit’ instruments the ‘daylight mode’ visualization (light background, dark numerals and needles) is inverted when it switches to the ‘night-time mode’ (dark background, illuminated numerals and needles). If we see this instrument either in broad daylight (lights OFF) and in complete darkness (lights ON) its design looks effective and contrast is good.

Unfortunately, this layout has some drawbacks in the ‘intermediate’ illuminance conditions that occur between night- and daylight, at dusk and dawn. (16) We must remember at this point that several countries, like Italy or the Scandinavian countries, mandate the permanent, 24-hours a day use of ‘Daytime Running Lights’ on highways and superhighways for safety reasons related to the improved visibility of the vehicles. (17) Figures reported by the Italian ‘Ministero dei Trasporti’ reckon that this safety measure, introduced in 2002, has been the main responsible for a sensible reduction of accidents, as we can see in this slide.  This success has been so impressive that EU is said to be studying the way to make the ‘Daytime Running Lights’ permanent on all new cars.

The instrument panel of an automobile is therefore illuminated most of the time. (18) This will not spoil the display’s readability in either extreme illuminance conditions, but when the ambient illuminance decreases, the luminance of the white/silver dial will fall sharply. (19) In certain moments it will be similar or identical to the luminance level of the backlit targets: as a consequence, there will be a condition of equiluminance and contrast will drop sharply.

When there is equiluminance between the background and target(s) of a given object that will be, in our case, the speedometer, its readability will become extremely difficult. Leibmann in 1927 remarked that “…an object that is demarcated from its surrounding exclusively by color looks less sharply defined than when the luminance of object and surround are not so precisely matched.” (20) Color contrast, or the idea of differentiating background and target(s) with a difference of color only, but without an appreciable difference in luminance between them, is therefore bound to give poor practical results: in fact, the border between two directly adjacent color fields looks blurred when there is no significant luminance contrast between the two colors. In other words, the resulting contrast is lower than the threshold of the ‘Minimally Distinct Border’. It is not our task here to see what hypothesis about the reduction of visual acuity at equiluminance is right, and what class of neurons is partially or totally silenced in this process: suffice to say that visual acuity decreases sharply with the fall of the luminance contrast.

(Click HERE to download the Power Point presentation of this paper) 

The circadian variations of illuminance during the 24 hours of the day will cause a constant change of the luminance of the surfaces in the interior of a car. This process is obviously slow and gradual, and it takes roughly 1 hour to complete a full cycle of ‘inversion’ of the dials’ contrast characteristics, from ‘day’ to ‘night’ mode, (21, 22, 23). This time will vary according to latitude, season of the year, geographical and meteorological conditions. For several hours of the day, especially at dusk and dawn, the luminance of a white/silver dial will be too close to the luminance of the numerals and needle to generate an appreciable luminance contrast. (24) In adverse weather conditions, for example in rainy or snowy weather with dark clouds, the illuminance can be so low that the dial will never receive enough light, even during the central hours of the day, to generate an appreciable contrast with the illuminated targets. As a consequence, for a long time the driver will be forced to try to extract data from an instrument where the contrast is much less than optimal for a good readability, until it becomes very low or almost zero.

Luminance contrast is one of the most important parameters of the readability of a display. Ophthalmologists know how many patients show a deficit in contrast sensitivity when in our practice we test it not only with normal central-acuity Snellen charts but with contrast-oriented test targets. A low-contrast target will require longer glances to be read correctly. Often, one glance will not be enough, so the driver will be forced to return to it again, wasting precious time in the process.

This is true for the average driver in good overall condition, physically healthy and with a good eyesight. The situation becomes even worse when a driver with impaired vision has to read those low-contrast instruments. As another author reminds us, “The minimum value of luminance contrast required to ensure display legibility is not a constant. It depends upon the size of symbols, observer age, background luminance, adaptation luminance, the amount of color contrast present and other contextual factors.”  Old age, opacification of the lens, corneal scars after infections or refractive surgery will worsen the contrast sensitivity of the driver, making a low-contrast display harder to read. For many drivers whose visual system already has a low contrast sensitivity, reading a low-contrast instrument is a difficult task. Moreover, the sensitivity of the eyes to contrast falls with the decrement of background luminance; on the other hand, older eyes are more sensitive to glare, so that the older driver will be annoyed by a brightly lit display and thus will turn its lights down, reducing the contrast even more.

Even a low-contrast display can be read if one can spend as much time as necessary in this task, but it would be dangerous to do so while the subject is driving an automobile. The main point is that reading the speedometer is necessary, but is still a secondary task and the primary task is always looking at the road ahead. We mustn’t design an interface so that a driver will be forced to spend a long time trying to read it. This would create the risk of diverting his eyes off the road for a dangerous amount of time, unavoidably increasing the driver’s mental workload and collision dangers.

Luminance contrast is too important for our visual system to compromise it for ‘style’. The visual acuity drop can be quite important in patients developing, for example, a corneal haze or a cataract. Senior citizens with a good BCVA will often have a significant diminution of the contrast sensitivity anyway, regardless of any pathological condition, just as a result of senescence; for them it will be time-consuming and sometimes altogether impossible to read a low-contrast dial that could still be read quickly by a younger driver.

Another event that will make reading a display even more difficult at dusk and by night is the increase of presbyopia that, according to Kuhl, will be of 50% when ambient illuminance falls to 0,1 cd/m² and 75% when it is below 0,01 cd/m². This worsens the reading of an object that is 60-80 cms from the driver’s eyes, as is a typical instrument panel. But if the speedometer is placed further away from the driver to lessen the effects of presbyopia, it could become unreadable for senior drivers wearing bifocal lenses.

In other words, when the ambient light decreases it will be even more difficult, for many drivers, to focalize the instrument panel; if this is not correctly designed, the strain of the driver’s visual system trying to obtain information from it will be significantly greater than average, and anyway undesirably high. It is enough to change the layout of the instruments from dark letters on white background to the classic white letters on black background to restore a perfect readability in any condition. (25)

All this means that the speedometer of some vehicles, for several hours of the day, is practically unreadable. This is not only undesirable: it’s not lawful, as we have already seen, and even in the countries where legislation is not explicitly stating anything about the readability of a speedometer, offering an accurate, easy-to-read instrument would have to be a top priority for any interior architect. (26) Ironically, nowadays it’s easier to find ergonomically efficient instruments on cheaper cars, where the budget probably do not allow flights of fancy and it is mandatory to fit the least expensive instrument available, that is the basic analogue white-on-black instrument which also happens, incidentally, to be the best one. It is equally puzzling to see that the white instruments fitted to expensive motor cars to offer more ‘sophistication’ to their customers are practically unreadable for long times of the day. So we have cheap, slow cars with excellent speedometers and very fast, expensive bolides with unreadable instrument panels. Is this rational?

An indirect confirmation of the drawbacks of this design comes from one of world’s more famous and more technologically-advanced sports-cars manufacturers, whose display engineers evidently recognized the functional limits of the white/silver dials installed, for ‘style’ reason, in their leading models. Though the white dials were retained, the engineers adopted a radical countermeasure, fitting an additional, large digital speedometer whose display is placed in a window under the existing analogue white dial. This digital display at least ensures the compliance of this automobile with the law as far as the speedometer readability is concerned, raising however some serious doubt about the rationality of all this, as an analogue speedo can be read in 0,6-0,7 seconds while a digital display typically takes 1,2 seconds. (Bhise, Forbes and Farber, cit. by Wierwille, 1986) (27) Please note that the ‘base’ version of this same automobile would not need this escamotage as it is fitted with perfectly readable black instruments.

The duplication of the speedometer on this car, understandable if one thinks of the possible legal implications of this situation, is an implicit admission of the ergonomic failure of a layout based on white dials.

 CONCLUSIONS 

Though one must accept certain aspects of innovation in the ever more competitive automotive market, we must at the same time admit that not all these innovations are ergonomically recommendable. An optimal readability in every illuminance condition should be a primary characteristic of any instrument, but the white dials used in some IPs do not guarantee this result because of the variation of contrast induced by the constant change of luminance of the dial and of its backlit numerals during the circadian variations of illuminance.

It would be advisable to remember that to achieve a maximum ergonomic efficiency, form (and, in this case, also color) must always follow function. A visual interface that is easy and quick to read is an absolute necessity, especially keeping in mind the potential dangers of a car out of control and the wide range of visual/mental capabilities of the subjects allowed to drive a motor vehicle. What is perfectly readable for a physically fit 20-years-old person could be a serious nuisance, and a cause for loss of attention, for a senior citizen.

From the point of view of the ophthalmologist, there is no reason to adopt this type of instrument having then to make amendments for its drawbacks, when an analogue instrument with black dial and white numerals fulfils admirably every possible need as far as the indication of the vehicle’s speed is concerned. Every in-car task, especially the secondary ones, must be engineered so that they do not disrupt the efficiency of the visual system, therefore adopting these ‘white dials’ does not appear to be advantageous. In fact, this layout may cause an undesirable elongation and multiplication of the in-car glances during the twilight hours, when its contrast is very low or near to zero and driving is already quite difficult without having to waste more time reading the speedometer.

Dott. Stefano Pasini MD, Bologna/Rochester, 25 June 2005