[tchkreddy]

Hi Experts,

While I understand from various forums that DTE is not very accurate due to lot of parameters impacting it, I'm curious as to how it gets calculated in first place. It is basically Fuel * Average Mileage.

I was trying to figure out which Average Mileage it picks from in my Kia Sonet HTK+. Since both 'Remaining Fuel' and 'Average Mileage' are both reported by Car only, I wanted to freeze one to discover another, so used full tank method.

Here are the numbers before refuelling.

DTE: 49
Accumulated Avg: 9.1 (Trip meter with manual reset)
Last Drive Info Avg: 18.2 (Trip meter reset on every ignition)
Fuel Economy Avg: 9.8 (Trip meter reset on refuelling)

I did a full tank with 40.74 L

Now DTE is 604. Means a jump of 555 km.
555/40.74 = 13.6 kmpl

Can anyone help me with how it derived 13.6?
Is it simple average of Accumulated Avg & Last Drive Avg? (18.2 + 9.1)/2 ?

[Rehaan]

Originally Posted by tchkreddy

Is it simple average of Accumulated Avg & Last Drive Avg? (18.2 + 9.1)/2 ?

The "algorithm" used for calculating DTE is bound to be more complex than a linear equation.

Tweaks to combat human psychology, as well as the manufacturer's liability (in terms of stranded drivers!) will be worked in.

Each manufacturer would have a different formula too -- which is why owners feel that certain cars have a much more reliable DTE than others.

In other words, its incredibly unlikely that you'll be able to reverse-engineer the formula, despite your best testing efforts.

[Dr.AD]

Originally Posted by tchkreddy

Is it simple average of Accumulated Avg & Last Drive Avg? (18.2 + 9.1)/2 ?

It is not based on long term ("accumulated average") at all, nor is it based on what is displayed in your "last drive average". Both of these numbers can be reset by the user any time and they will vary based on when the last rest was done (i.e., since when are you accumulating these averages).

Instead, in most cars, the Distance to Empty (DTE) is based on most recent mileage as computed by the car. Most cars use something like last 30 minutes or last 30km (or some combination thereof) worth of fuel consumption data and extrapolate that to the remaining fuel quantity to "estimate" the DTE.

This is also why the DTE can vary noticeably when you drive on a smooth highway for about 30km or get trapped in congested city traffic. Whereas the long term average changes very little based on these immediate conditions.

Originally Posted by Rehaan

The "algorithm" used for calculating DTE is bound to be more complex than a linear equation.

Exactly! Even what I explained above is an oversimplified logic. In reality these calculations are done by ECU in real-time based on instantaneous measurements and then integrated over time. And on top of that, there are some "correction factors", which are almost always empirical, applied. The resultant formula is not the simple linear average.

Each manufacturer would have a different formula too -- which is why owners feel that certain cars have a much more reliable DTE than others.

Very true! Once my company was in discussion with developing this logic for an electric scooter OEM in India. They wanted something very specific to them based on their own requirements and user study. There is no common formula that all OEMs use.

[Mustang_Boss]

The Distance to Empty shows how much more you can drive at your current pace.

What i could understand is, [from Maruti cars]

- If you are driving in 1st gear, in Bumper to bumper traffic, It would show as say 100.
- The moment traffic starts moving, and you switch to 5th gear and at 50kmph, say, it would show as 220 or something.

It calculates the DTE based on
- Amount of Fuel available in the tank. (Till Last point, not considering the reserve)
- What's your current consumption, since last three/five checks (varies by car). These checks will be periodical, like for every 30 seconds or 1 minute.. The Chips used in the car will determine the accuracy. If they used good chips, they could do complex calculations. In basic cars, the calculations would be much simple.

[audioholic]

Each ECM supplier will have their own way of implementing Range or DTE of a given car. They have different weights given to historical average, current average FE, current average speed and so on and each of them would have aimed to achieve an accurate estimation but while remaining on the safer side. That is, most of the time you can outrun the range but not fall short of it unless in extreme conditions.

Typically the range = Amount of fuel left * Avg FE(Weighted)

Avg FE Is a weighted average of FE of trip, since refuel and long term FE since manual reset.

Increase in the avg FE over time due to betterment in the driving conditions is always combined with reduction in the fuel level in the same time period. Hence the range always would just reduce but at a lower pace even if you might have covered more distance. This happens in my Slavia and Celerio. But when the inverse happens, avg FE drops along with drop in fuel level which might lead to Range depleting faster than you expect.

The only time you see a major jump is while refueling. Thats when one of the components of the avg FE ie FE since refuel will change. This will be reset to zero and when zero, there will be an assumed value taken for the calculation. Hence there will be a jump in range in case just before the refuel there has been above average economy.

In the Innova it is most accurate as per my experience amongst my other cars and Celerio least accurate. Slavia is on a safe side all the time with lesser range shown and if I drive extremely well, the range remains constant.

[Dr.AD]

Originally Posted by audioholic

Increase in the avg FE over time due to betterment in the driving conditions is always combined with reduction in the fuel level in the same time period. Hence the range always would just reduce but at a lower pace even if you might have covered more distance.

The only time you see a major jump is while refueling.

I agree with you on most points. However, even this implementation is very car specific. For example, in my BMW 320d, there are many situations other than refueling when the DTE actually increases.

This is because BMW calculations seem to give a very heavy weightage to last 30 minutes of driving. Therefore, if I drive in the city traffic for a while (when the FE is low and the DTE is lower), and then drive onto an open highway, the FE increases and then the DTE increases significantly. This happens to me all the time in my outstation drives. I start with a low DTE while navigating city traffic till I reach the highways, and once I start on the highway, I see the DTE number climb up nicely and consistently till a point, and then once it reaches a peak corresponding to my last 30km of highway FE, then it starts dropping linearly as the fuel level goes down. This happens without any refueling in-between.

There are many occasions where I have seen the DTE in my BMW 320d increase significantly as I drive on better conditions than previous hour or so, without refueling!

In my Maruti SX4, I had installed an Android Headunit and set up my own dashboard using Torque Pro app. There, I had set all sorts of custom equations and custom parameters. One of the parameters was DTE. There I had set a formula that was a combination of long term FE and the FE since the last fueling. There I observed the DTE behavior exactly as you mentioned. But certainly not in BMW, which has a different way of calculating DTE.

[electro-monk]

Originally Posted by audioholic

Increase in the avg FE over time due to betterment in the driving conditions is always combined with reduction in the fuel level in the same time period. Hence the range always would just reduce but at a lower pace even if you might have covered more distance. This happens in my Slavia and Celerio.

This may not apply to petrol Swift.
For instance, if I have to go towards Krishnagiri on Bangalore-Kanyakumari highway after refuelling at the start, I have noticed that the DTE stays same at 710 for the first 30 kms and then after I touch the 6 laned NH44, it gradually increases to 750. This max reading comes around Shoolagiri, after which it starts declining. Since Shoolagiri is 60 km after refueling, so I find that total range is 750+60 = 810kms on full tank for my car.

[Auto_guy_101]

Originally Posted by tchkreddy

Hi Experts,

While I understand from various forums that DTE is not very accurate due to lot of parameters impacting it, I'm curious as to how it gets calculated in first place. It is basically Fuel * Average Mileage.

This is a rather interesting topic. Some information I noticed on my drives and conclusions I have are
1. As many have pointed out, the calculations are extrapolations of the current driving conditions and can't exactly be reverse engineered.
2. The lower processing power of vehicle chips (say compare to a smartphone) and type of sensors (esp in the non premium category, I guess. I'm no expert) leads to considerable variations like below
3. Say you're on a highway cruising at 80-110, and fill up fuel (not necessarily a full tank even), some cars have a slight gradual increase in dte ( say 200 -> 450 (after filling) - > 500 over some time. My best guess is, in cruising speeds the instantaneous fuel consumption isn't mentioned correctly.
4. Pt 3 is more pronounced if you fill the tank.
5. Unless you have a considerable regular commute (say 20+ km one way), with limited traffic, the tank full mileage for mixed highway+ city would be pretty meaningless financially.
Most cars with a low daily commute crunch their miles on open roads. Meaning, for a large part of your travel, you're sipping fuel but once in the city traffic, you're guzzling it. So it would be more ideal to try a tankful mileage for the open roads alone

[MT_Hyderabad]

As per my experience in a Maruti Ciaz, this is how it is calculated.

The fuel sensor knows how much fuel you have(before the low fuel light comes up)
The throttle input sensors also know how you are driving (current mileage).
It does simple mathematical calculation to ascertain how long you will be able to drive before the low fuel warning comes up.

If I accelerate rapidly, I have seen it go down from 66 to 35! in just 15 seconds gap.

[Bonemechanic]

I drive a Marazzo M8 and I reset my trip meter to zero on the first of every month. The Marazzo's Driver Information Display shows intervals of 25 km from the start of the drive, such as 25 km, 50 km, 75 km, and so on.
The Ecosense app in the ICE also displays the distance travelled and fuel efficiency (FE), among other things. I have noticed a difference in the distance travelled displayed on the MID (Multi-Information Display) and the Ecosense app.

The upper part shows the MID trip B reset to zero, while the lower part displays the Ecosense reading at the same time.


The upper part shows the MID reading at the end of the trip, while the lower part displays the Ecosense reading.


From the pictures, it is evident that there is a disparity of 900 meters. The Ecosense app shows an approximate difference of 1 km in a 45 km drive.
I have driven approximately 118,000 km, so there is approximately a 2,400 km difference between the odometer reading and the Ecosense reading.

Regarding the second issue, the distance from my home to the office is approximately 27.2 km, according to the MID. There is a display on the MID that appears after completing 25 km. I am aware of this point, as it is only 2.2 km away from my destination, which is also confirmed by Google Maps. However, after changing to new tires, I noticed that the 25 km mark on the MID has shifted approximately 100 meters towards the destination, resulting in a slightly shorter distance. This difference is also confirmed by Google Maps.
A is home
B is previous 25 km mark
C is new 25 km mark
D is Hospital parking slot


My doubts are
Why there is a differance in MID and ecosense reading?
How the distance is reduced after tire change?

Thank you, @libranof1987, for guiding me to this thread.

[singhals10]

I used to think that it is based on ARAI certified fuel effeciency of the car, and better you drive - closer to ARAI based DTE you will achieve.
Or atleast some weightage is given to ARAI fuel effeciency.
Feel free to inform your opinion on the above.