As a User Experience Designer and a car lover, it was only natural for me to take a closer look at the Tesla Model S instrument cluster, and more specifically the speedometer and the energy indicator. Here is what I found.
The current design of the speedometer/energy indicator is shown in the image below. Note that this picture is taken from an 85 kWh car.
Take a look at the left part of the dial, the speedometer part. One interesting point to note is that the scale Tesla has used for the speed is not constant. Speed is indicated in 10 mph increments with a thin or thick bar, with numbers at every 20 mph. What is very interesting is that these intervals are not of the same length!
I overlaid a 24-slice circle on the dial to make it easier to see.
From 0 to 60 mph, the intervals fall along the 24-slice guide, but from 60 mph on, they no longer match up. The distance between 60 mph and 80 mph is a little bit shorter than the distance between 40 mph and 60 mph.
Also, the distance between 80 mph and 100 mph is shorter than that between 60 mph and 80 mph. So from 60 mph on, the intervals also get smaller and smaller.
In addition, there is no visual indication to the driver of this fact.
The problem with this shrinking scale is that it gives the driver an incorrect impression about his acceleration.
Imagine you are driving the Tesla Model S, merging onto a highway, and you are accelerating at a constant rate from 0 to 75 mph. You will see your speedometer needle go from 0 to 60 mph at a constant speed, telling you that your acceleration is constant. Then, for no apparent reason, above 60 mph, the speedometer needle will actually decelerate until you reach 75 mph!
Consciously, the driver can look at the big white speed number and see that the car keeps accelerating at the same rate, but the impression he is getting from his subconscious or reptilian mind will be different. This causes confusion.
Why did Tesla do that?
Well, I can only guess as to what they were thinking, but these are my two assumptions.
Firstly, since the top speed of the Model S is around 120 mph, using a constant scale would have meant that the 60 mph mark would have been located at the 9 o’clock position; this would have meant that under 60 mph, the needle movement would have been restricted to the first half of the left part of the dial. Since most of our driving is done between 0 and 60 mph, that would not have been a very efficient use of space.
Secondly, and more importantly, it would have broken the 24-slice grid that Tesla has used for this dial. Every good design starts with a good structure, and in Tesla’s case that structure is a 24-slice circle.
The 0 mph mark starts on a vector; the 10, 20, 30 mph and so on each fall on a vertex; the 0 kW starts on a vertex. Tesla clearly wanted the scale of the speedometer to fall on the 24-slice grid.
So, Tesla decided to follow the 24-slice grid up to the 60 mph mark to give a larger area to speeds up to 60 mph, but they still had to fit another 60 mph's worth of increments from there to the top of the dial, and so they had to shrink them more and more up to the 120 mph mark.
An alternate solution
Instead of keeping the intervals constant to 60 mph, why not follow it up to 80 mph? Since most people (I assume) in the United States rarely go over 80 mph, they would never suffer from the deceptive needle speed.
Then, for people who really want to go at “unlawful speeds (in the US)”, why not flip to a different scale, but one that would nonetheless remain constant from 80 to 120 mph (or whatever the actual speed limit is)?
This alternative would fully respect the 24-slice grid and would have 2 scales: one for speeds under 80 mph (where most people drive) and one over 80 mph (which I would call warp speed!). The "swish" between the 80 to 120 mph marks would indicate the fact that the scale has changed.
In conclusion, this alternate design:
- contains two constant scales instead of a changing scale;
- notifies drivers of the change in scale with the swish; and
- means that drivers driving under 80 mph (most people) will never encounter a changing scale.
Let’s now take a look at the right side of the dial, the energy indicator. As we’ll see, it suffers from some of the same problems as the speedometer. The image is of the 60 kWh Model S. The upper part shows the instantaneous energy consumption and the bottom part shows the instantaneous energy regeneration. The picture at right currently shows the car consuming about 15 kW.
Since you have now been trained in looking at scales, you will readily notice that the consumption scale (the upper part) is not linear either. It is numbered as shown:
15, 30, 60, 120, 240
The first interval is worth 15 kW and then it doubles per interval; it is a geometric scale. The same interval is used for energy regeneration, which is a good thing. We want to have the same scale for consumption and regeneration so that when the driver accelerates at 60 kW and then lifts off the acceleration pedal and regenerates at 60 kW, the trailing needle will have the same length to show that the same amount of energy was consumed and then recuperated, minus friction.
The problem I have with the geometric scale this time is that it is too sensitive. I have yet to drive a Tesla myself, but in watching people on Youtube filming themselves driving Teslas, I've noticed that as soon as they start driving, the consumption needle goes halfway up the dial; remember, this is a big car and it takes a lot of energy to get it going. So once again the driver is getting the wrong impression, namely, that accelerating at 60 kW represents approximately 50% of the total kW output of the car. This is wrong, because 60 kW out of 240 kW represents 25% of the possible output of a 60 kWh Model S.
Drivers are then getting the impression that the Model S consumes much more than it actually does and, conversely, that it regenerates much more than it actually does. A 60 kW regeneration is still only a quarter of what the car can really output (240 kW), not half of it as based on the needle length.
I understand Tesla had to use the same scale for both consumption and regeneration; since the regeneration scales had to use the remaining space down to the gear selector, they had no choice but to go that route.
The second problem
There is an additional problem, this time, with the display of the 85 kWh battery car (image at right).
Because the energy consumption needs to go to 320 kW instead of 240 kW and still maintain the geometric scale, Tesla had to make the consumption scale different than the regeneration scale. This means that accelerating at 40 kW will show a trailing orange needle with a length of 2 intervals and the corresponding 2 intervals of regeneration will be equivalent to only 30 kW, again giving the wrong information to the driver’s reptilian brain.
The solution (which you'll have guessed if you are still reading this “short article”) is to change the scale to a linear scale in order to respect:
- the 24-slice grid;
- the linear scale for the reptilian brain; and
- the fact that the 60 kWh car can output around 240 kW.
We get this display with a 50 kW increment that goes up to 250 kW.
Since the regeneration scale needs to be the same as the consumption scale, we end up with a much smaller regeneration area, which is a perfect representation of reality. Regenerating power recuperates only roughly 20-25% of the maximum output of the battery.
We can solve the second problem for the 85 kWh car the same way,
but now using a 60 kW increment scale that goes up to 300 kW.
In both of these solutions, for the 60 kWh car and the 85 kWh car, the maximum kWh number shown on my scales does not represent exactly the maximum output the car is capable of. I think this is a small price to pay in order to have linear scales for the energy indicator.
I created a new problem
The problem I've created is... empty space between the gear selector (for lack of a better term) and the end of the regeneration scale.
However, free space is not really a problem in interface design - it actually represents opportunity.
So what additional information can we give the driver in that space?
Reading the Tesla forums, I realized that some drivers did not understand how the battery pack temperature can impact both the maximum energy the car can output and the amount of regeneration the car can take.
Tesla notifies the driver that the battery pack is too hot by showing a dotted yellow line limiting the maximum energy the car can consume. When the battery pack is too cold, the same yellow line caps regeneration to an amount lower than 50 or
Battery pack temperature is important and drivers who have a cold battery pack are not made aware of how fast the pack is reheating while driving, so I think it would be nice to show them.
The white line represents the current battery pack temperature and the two yellow zones would show when the pack is either too cold (at the bottom) or too hot (at the top). Watching the white needle move through time would give the driver an idea of when the pack would go back to normal temperature.
This temperature display would only be shown when the battery pack is outside normal temperature range and would be hidden when the battery pack is at normal operating temperature.
In conclusion, unless there is something I am missing here, Tesla’s use of variable scales (in the case of the speedometer) or geometric scales (in the case of the energy indicator) is giving Tesla Model S drivers the wrong impression.
Switching to linear scales fixes these problems, and in the case of the energy indicator, also provides a new opportunity to provide some useful information to the driver.
I would really like to hear comments from Tesla drivers who have experienced the problems I explored in this post.
And I would also love to hear from Tesla to explain the reasoning behind these design decisions.