First time user of ABRP - trip planning question

[darth_vad3r]

I believe you’re right, but I’m not sure of the locations that data was generated from. If 70% of the vehicles were located in California generating the data, winter time might be way off pace with the model. SoCal has pretty mild winters compared to a lot of places... unless the data is weighted, which it might be in the model, the further away from that California weather the more inaccurate the model will get...

Yes, there definitely could be location bias. I think they have people contributing from Norway and California though ... I think they also do adjust the data gathered for elevation changes and temperature somewhat though ... and they do allow you to input a temperature for the prediction, implying the model is based on a model temp.

I've found it to be remarkably accurate for me when I've used it with a realistic boost on the reference speed to match my driving. I mean within 1-2% for a trip that is going through elevation changes and hundreds of km... considering Tesla has 'insider' info for their nav screen charts and estimates ... ABRP has done a great job.

 

[acarney]

Yes, there definitely could be location bias. I think they have people contributing from Norway and California though ... I think they also do adjust the data gathered for elevation changes and temperature somewhat though ... and they do allow you to input a temperature for the prediction, implying the model is based on a model temp.

I've found it to be remarkably accurate for me when I've used it with a realistic boost on the reference speed to match my driving. I mean within 1-2% for a trip that is going through elevation changes and hundreds of km... considering Tesla has 'insider' info for their nav screen charts and estimates ... ABRP has done a great job.

I think the temperature input doesn’t factor in HVAC usage but instead factors in reduced energy from cold packs and stuff like that. I think... I’m always surprised going from 95F to 25F makes a very small difference... but I know running heat makes a huge difference in the real world...

 

[darth_vad3r]

I think the temperature input doesn’t factor in HVAC usage but instead factors in reduced energy from cold packs and stuff like that. I think... I’m always surprised going from 95F to 25F makes a very small difference... but I know running heat makes a huge difference in the real world...

Yes, e.g. if you raise the temp from 20C to 30C it gets a bit more efficient (in real life, and in ABRP), but ABRP doesn't try to factor in that you will be using AC, and likewise for colder temps, as you noted, it doesn't count any heating.

I guess you could be wearing a parka and not using heat ... or just the seat heater ... so ya, if you are making cold-weather trips with heat, you need to add a fudge factor in somewhere to account for that--increase reference consumption or reference speed maybe?

 

[Rocky_H]

I think the temperature input doesn’t factor in HVAC usage but instead factors in reduced energy from cold packs and stuff like that. I think... I’m always surprised going from 95F to 25F makes a very small difference... but I know running heat makes a huge difference in the real world...

What?! Is ABRP really that pathetic that it wouldn't count HVAC use? I just went to check the settings, and WOW! You're right! It doesn't let you change that. Well that's yet another reason I will never use ABRP. The EVTripplanner site has two separate settings for "inside temperature" and "outside temperature" so that it does model heating and air conditioning use.

 

[darth_vad3r]

Agree with all this, except ... percent is a lot LESS meaningful.

Percent is relative to "100%" ... and "100%" can change ... from degradation, miscalibration, software-nerfing by the mothership, impending battery failure, etc.

If you don't look at 'miles' you won't realize if your 100% is dropping fast because of some real issue (other than the typical 3-5 miles people run around with their heads cut off aboute).

If you use miles on the dash, you can see % on the trip energy chart at the same time.

3 people disagree with the facts posted above. #FakeNews. LOL.

Every single word above is a true statement. What part are people disagreeing with? Anyone?

 

[M3BlueGeorgia]

Could it be that it recalculates the arrival estimate continually and so it will tend to get more and more accurate as you get closer to the destination? A smaller distance remaining means a smaller error typically if the error is percent-based.

There is some truth in that, but it shouldn't have much effect over the first (say) 30 miles of a 200 mile leg.
I had an extreme example last week where the forecast arrival percentage increased from 8% to 24% over the first 30 (or so) miles of a long leg.

 

[darth_vad3r]

There is some truth in that, but it shouldn't have much effect over the first (say) 30 miles of a 200 mile leg.
I had an extreme example last week where the forecast arrival percentage increased from 8% to 24% over the first 30 (or so) miles of a long leg.

That is pretty extreme. Did you follow the nav directions? I notice it recalculates silently (and better) when I ignore the way it wants me to go and go a flatter way. Also if I go slow up a hill it will improve a bit. I'm only talking 1-2% here either way, but this is a short ~25-mile commute.

 

[M3BlueGeorgia]

That is pretty extreme. Did you follow the nav directions? I notice it recalculates silently (and better) when I ignore the way it wants me to go and go a flatter way. Also if I go slow up a hill it will improve a bit. I'm only talking 1-2% here either way, but this is a short ~25-mile commute.

It was on a long trip, on the way back from NYS to Atlanta on I-81. Basically we were stopping every 200 miles (or so) to recharge and attend to personal issues.

The daytime temps were in the low 90s(F) and there was a short period of time when the car was driving at low speed and then just parked and cooling down the interior, so had a weird battery usage peak. We had taken a meal break, so I had to go out to the Supercharger during the meal to move to car to the restaurant because we had charged enough for the next leg.

 

[edigest]

@edigest , @KenC is calling you a straw man figment of my imagination.

After reading over the last couple days of this thread I have to agree with @KenC that you're using a straw man argument. You decide how you want to define the issue and the build your argument around it. Voilà, you can't be wrong!

You said, "Miles is more informative because it INFORMS you of how much energy is in the battery." Actually, it doesn't. Miles is a representation of the SoC and application of a formula that factors in other assumptions about consumption.

Here's a simple experiment. I charge in the evening when the temperature is, say, 100 degrees. When charging is complete, the miles show 217. Next morning, when it is 80 degrees, the miles shows 220. Did energy in the batter increase overnight while the car was sleeping?

You're very passionate about your way. That's fine. Plan around miles and have lunch at Burger King.

 

[darth_vad3r]

After reading over the last couple days of this thread I have to agree with @KenC that you're using a straw man argument. You decide how you want to define the issue and the build your argument around it. Voilà, you can't be wrong!

You said, "Miles is more informative because it INFORMS you of how much energy is in the battery." Actually, it doesn't. Miles is a representation of the SoC and application of a formula that factors in other assumptions about consumption.

Here's a simple experiment. I charge in the evening when the temperature is, say, 100 degrees. When charging is complete, the miles show 217. Next morning, when it is 80 degrees, the miles shows 220. Did energy in the batter increase overnight while the car was sleeping?

You're very passionate about your way. That's fine. Plan around miles and have lunch at Burger King.

LOL. Everything in the car is an estimate in one way or another. Estimates fluctuate based on inputs, including temperature.

At any given time, the BMS's current estimate of kWh of ENERGY available in the battery is displayed to the user, e.g. on an LR AWD Model 3 by dividing the kWh by a constant, such as 245 Wh/mi.

So if the car estimates 53.17 kWh of energy is in the battery in the evening, it divides it by 245 Wh/mi and arrives at a user-facing representation of that amount of energy of 217 miles.

In the morning if it re-estimates that the energy available in the battery is now 53.9 kWh, it divides by the same 245 Wh/mi to calculated 220 miles and show that to the user.

If you have the original 310 miles of range, these numbers will show as 70% or 71%.

If your battery has degraded 10% and only has 279 miles of range, these numbers would show as 78 or 79%.

Now if you'd set your charge target slider to "70%" in the app, on the non-degraded 310-mile capacity car, you would have charged to 217 miles at night, and woken up to 220 miles.

In the 10%-degraded 279-mile capacity car, you would have charged to 70% and woken up to 71% and thought everything was hunky-dory... except... if you looked at the MILES you would be *MORE INFORMED* about the state of the car and realize that your 70% charge target somehow only got you to 195 miles!

195 miles vs 217 miles is *MORE INFORMATION* than "70%" vs "70%".

Miles informs you MORE than percent does. Period.

You said, "Miles is more informative because it INFORMS you of how much energy is in the battery." Actually, it doesn't.

so, actually it does. See above. 217 miles or 195 miles is more information than "70%" or "70%".

Miles is a representation of the SoC and application of a formula that factors in other assumptions about consumption.

Wrong, miles (on the battery gauge, instead of %) is a representation of the estimate of ENERGY (kWh) in the battery.
Percentage is a representation of that estimate (kWh), divided by a SECOND estimate (kWh) of the total current maximum capacity of the battery (which may be less than day 1 because of degradation or other factors).

If you only look at % you won't know what the 2nd number is. This is less informative.

 

[Phlier]

Percentage is a representation of that estimate (kWh), divided by a SECOND estimate (kWh) of the total current maximum capacity of the battery

That right there really drives it home... at least, for my brain.

I really appreciate all the hard work you guys have put into the posts in this thread. Especially our Sith friend.

Thanks for educating us... much appreciated.

 

[darth_vad3r]

That right there really drives it home... at least, for my brain.

I really appreciate all the hard work you guys have put into the posts in this thread. Especially our Sith friend.

Thanks for educating us... much appreciated.

Any time.

So maybe this is the key point we should emphasize. There are *two* estimates, and one constant (based on model and EPA rating).

1. E1 is the current *estimate* of energy (kWh) in the battery.
2. E2 is the current *estimate* of maximum energy capacity (kWh) of the battery (can be less than new because of degradation, etc)
3. C is a constant for your model/trim. (e.g. (1/(245 Wh/mi) for LR AWD)

• Miles is: E1 * C
• Percent is: E1 / E2

(nobody tells you what E2 actually is, you can estimate it by doing: C * "miles" / "percent"

If you just want to know how much energy is in your battery, so you know how far you can drive, Miles will tell you this most directly, because E1 is the answer, and Miles is just E1 multiplied by a constant so our brain can read 'rated miles' which we understand a bit better than "kWh".

If you use percent, you know how close to E2 you are. E2 is your current "full" estimate. If you know what E2 is, you can get as much information from 'Percent' as Miles. They don't tell us what E2 is though... unless maybe you are monitoring the CAN bus.

You could convert "Miles" to "original percent" by dividing it by your rated miles (say 310). If you did this, you'd see how the two percent numbers can be *different*, thus the miles one is giving you the true state of the battery in terms of how much energy you have to use to drive. When new these two percent numbers "percent of original" and "dashboard percent" will be the same. As you use the car different factors will make "dashboard percent" drift due to real issues like degradation and transient issues like calibration drift.

If you always use "dashboard percent" you can't see these drifts happening.


ps. This right here made me chuckle ... the "Informative" tag next to the last words in the post

 

[Phlier]

Any time.

So maybe this is the key point we should emphasize. There are *two* estimates, and one constant (based on model and EPA rating).

1. E1 is the current *estimate* of energy (kWh) in the battery.
2. E2 is the current *estimate* of maximum energy capacity (kWh) of the battery (can be less than new because of degradation, etc)
3. C is a constant for your model/trim. (e.g. (1/(245 Wh/mi) for LR AWD)

• Miles is: E1 * C
• Percent is: E1 / E2

(nobody tells you what E2 actually is, you can estimate it by doing: C * "miles" / "percent"

If you just want to know how much energy is in your battery, so you know how far you can drive, Miles will tell you this most directly, because E1 is the answer, and Miles is just E1 multiplied by a constant so our brain can read 'rated miles' which we understand a bit better than "kWh".

If you use percent, you know how close to E2 you are. E2 is your current "full" estimate. If you know what E2 is, you can get as much information from 'Percent' as Miles. They don't tell us what E2 is though... unless maybe you are monitoring the CAN bus.

You could convert "Miles" to "original percent" by dividing it by your rated miles (say 310). If you did this, you'd see how the two percent numbers can be *different*, thus the miles one is giving you the true state of the battery in terms of how much energy you have to use to drive.


ps. This right here made me chuckle ... the "Informative" tag next to the last words in the post

View attachment 454163

Perfect!

That right there needs to be made into its own thread and stickied. Once you get this concept, it just makes sense. I really wish they would just tell us what the actual current and max kWh is/are, instead of turning it into a "miles" estimate. At least we can figure it out now that you've spelled it out. Giving us the ability to see both their miles estimate and the actual current energy capacity (kWh) would be very nice.

And yes, the fact that it is less informative is very informative.

 

[darth_vad3r]

@M3BlueGeorgia ,really? Still disagree? At this point I feel like you are just trolling me with 'disagrees', or refusing to agree with something like "1+1=2".

Please explain in a short sentence or two what point you are disagreeing with, because I think there is just a misunderstanding underlying your disagreement. Unless you are actually trolling. In which case, bravo.

 

[M3BlueGeorgia]

Unfortunately the miles presented next to the current speed is just the theoretical EPA range and takes no account of your specific tire choice, nor your driving patterns, nor the weather conditions, nor the speed you are about to drive, nor the road conditions.
Often its just a bit wrong (which is OK), other times it can be dangerously wrong.

The Navigator, when it estimates your arrival percentage, takes into account recent battery usage, and therefore its forecast gets better once you start into your trip because it keeps dynamically recalculating the estimated arrival percentage. I think it also takes account of future traffic on your trip, but I'm not sure about this.

I quit showing the miles and switched to percent last winter.
I was driving in the MidWest with daytime temps in the teens and low 20s (F), and the EPA miles were massively wrong because the temperatures drove my range down by 40+%. The higher (80mph) speed limits in Kansas didn't help either.
If you spend all your driving in Florida or Southern California, it probably doesn't matter whether you show miles or percent, as the EPA miles won't be too wrong, but if you encounter mountains or severe cold, the EPA miles can be way off.

@darth_vad3r: Noting that you appear to be a new-ish owner in Canada, it might be that you'll change your mind also come December

 

[Phlier]

In which case, bravo

10/10 would troll again.

 

[Phlier]

I think it also takes account of future traffic on your trip, but I'm not sure about this.

Oh man, that reminds me... I've been meaning to ask about this, as well. I've noticed exactly the same thing happen re: updating arrival percentage based on changing traffic conditions.

Here's what my situation was when I noticed this...

I'm in Las Vegas, desperately trying to find the supercharger shown on the map. Unfortunately, it appears that it's hidden deep within the parking structure of a large resort, and I just can't find the darn thing.

So I update my destination to a supercharger waaaay far away (right at the southwest airport boundary of McCarran airport for those interested). Nav says I'll have 23% when I get there. A few minutes later, I notice a big red line appear (bad traffic) along my route, and the nav updates the arrival percentage estimate to 18%. The route itself doesn't change. I ended up having 17% when I got there. Traffic delayed me by 40 minutes. Outside air temp: 119F.

 

[darth_vad3r]

Unfortunately the miles presented next to the current speed is just the theoretical EPA range and takes no account of your specific tire choice, nor your driving patterns, nor the weather conditions, nor the speed you are about to drive, nor the road conditions.

Right, it's a measure of the amount of ENERGY in your battery. It's not a unit of distance.

Often its just a bit wrong (which is OK), other times it can be dangerously wrong.

No, it's always 'right' in how much energy you have in your battery (barring estimation errors). It is *NOT* estimating how far you can drive. NOT NOT NOT. Use the navigation for this. It's estimating how much ENERGY is in your battery.

It's not road miles, it's 'rated miles' which is not a unit of distance, but of energy.

You are debating a different point altogether. The discussion was about if percent is more *informative* than miles. It is, because it gives you the most *information* about *energy* available to you in the battery.

@darth_vad3r: Noting that you appear to be a new-ish owner in Canada, it might be that you'll change your mind also come December

In the winter, I will find my X-mile commute will take 1.5x or whatever 'miles' extra off the dashboard (might be 1.3, might be 1.8 or 2x)
If I used percent, I would be doing the same mental math, 15% in winter instead of 10% in summer, right? Or 18% or 20%.

This is "equal" amount of information. .... except ....

Except the percent is actually LESS informative, because if I have battery degradation, the "15%" in 6 months from now might be LESS energy than the "15%" from right now, and less again 6 months after that. After a full year I might not make that bad winter commute 2x with "30%" and think I'm just driving worse or what's wrong? Why does it take 35% today. Must be colder? If I checked miles I'd see that ya, it must be colder, I'm using more miles. Or no, wait, I'm using the same miles but my 30% amount is a way smaller number than it was a year ago. Hmm.

"Miles" will always be the same X amount of energy in kWh, whether it is winter or summer. Of course your driving style, speed, weather will use more or less energy for a trip. This is exactly the same thing it does with "%" though.

That's a personal preference thing. You can debate that for centuries.

You can't debate the fact that miles tells you how much energy is left in your battery BETTER than dashboard-percent, because dashboard-percent can *CHANGE* over time, months, years ... days if you have rapid degradation issues!

 

[M3BlueGeorgia]

Right, it's a measure of the amount of ENERGY in your battery. It's not a unit of distance.



No, it's always 'right' in how much energy you have in your battery (barring estimation errors). It is *NOT* estimating how far you can drive. NOT NOT NOT. Use the navigation for this. It's estimating how much ENERGY is in your battery.

It's not road miles, it's 'rated miles' which is not a unit of distance, but of energy.

You are debating a different point altogether. The discussion was about if percent is more *informative* than miles. It is, because it gives you the most *information* about *energy* available to you in the battery.



In the winter, I will find my X-mile commute will take 1.5x or whatever 'miles' extra off the dashboard (might be 1.3, might be 1.8 or 2x)
If I used percent, I would be doing the same mental math, 15% in winter instead of 10% in summer, right? Or 18% or 20%.

This is "equal" amount of information. .... except ....

Except the percent is actually LESS informative, because if I have battery degradation, the "15%" in 6 months from now might be LESS energy than the "15%" from right now, and less again 6 months after that. After a full year I might not make that bad winter commute 2x with "30%" and think I'm just driving worse or what's wrong? Why does it take 35% today. Must be colder? If I checked miles I'd see that ya, it must be colder, I'm using more miles. Or no, wait, I'm using the same miles but my 30% amount is a way smaller number than it was a year ago. Hmm.

"Miles" will always be the same X amount of energy in kWh, whether it is winter or summer. Of course your driving style, speed, weather will use more or less energy for a trip. This is exactly the same thing it does with "%" though.

That's a personal preference thing. You can debate that for centuries.

You can't debate the fact that miles tells you how much energy is left in your battery BETTER than dashboard-percent, because dashboard-percent can *CHANGE* over time, months, years ... days if you have rapid degradation issues!

I get your point.

My solution is to occasionally briefly flip the display to distance when the car has just charged to 90% to see how things are going with the battery (and extrapolate to 100%). Currently after 27K miles (43km) I'm at about 304 instead of 310, so that's pretty good.
Then flip back to percent before driving. I probably do that every month or so.

 

[darth_vad3r]

Let's go back to the start of this.

I find ABRP very useful for a trip planning tool, because it shows Superchargers other than the one's recommended by the in-car navigator.
On a long trip I will sometimes select different Superchargers than the in-car navigator recommends, and ABRP is a good tool for planning to do that.

Couple of issue with your analysis:

1) Suggest you switch your standard battery status display to percentage instead of miles. The miles it shows isn't the range you are going to get. Percent is a lot more meaningful.

2) Best never to plan on an arrival percentage less than 10%.

3) Make sure the ABRP parameters are set correctly, otherwise it won't give you the correct forecast.

4) Seems like ABRP gave a pretty good forecast.

5) Once you were 30 miles into the trip, what was the in-car navigator forecasting for your arrival percentage?
Note: Because it uses recent driving behavior for its forecast, the forecast takes a while to stabilize to your current reality.

ABRP = abetterrouteplanner.com

I said I agreed with everything you wrote, except that "Percent is a lot more meaningful".

If you see, that means I agree that miles "isn't the range you are going to get".
Percent is also not the range you are going to get either though.

How is percent *more* meaningful? You perhaps meant a different word.

Perhaps you feel "miles" is more misleading, if you get misled into thinking it is the range you will get (it isn't the range you will get though). So if you just understand this point, that it's telling you energy left in the battery, not miles you can drive, then you can see why it has more meaning than "dashboard percent" because "dashboard percent" is how close you are to a number you don't know (dashboard 100% changes meaning. If it changes meaning how can it be more MEANINGFUL).

This is what I tried to express in my first response:

Agree with all this, except ... percent is a lot LESS meaningful.

Percent is relative to "100%" ... and "100%" can change ... from degradation, miscalibration, software-nerfing by the mothership, impending battery failure, etc.

If you don't look at 'miles' you won't realize if your 100% is dropping fast because of some real issue (other than the typical 3-5 miles people run around with their heads cut off aboute).

If you use miles on the dash, you can see % on the trip energy chart at the same time.

Then you disagreed with that.

I think what you are actually disagreeing with isn't what I'm saying, but on the definition of "meaningful"

mean·ing·ful
/ˈmēniNGfəl/
adjective

1. having meaning.
   synonyms: significant, relevant, important, consequential, material, telling, pithy, weighty, valid,worthwhile, purposeful
2. having a serious, important, or useful quality or purpose.
3. communicating something that is not directly expressed.

Dashboard-Miles communicates to you something that is not directly expressed, the amount of energy in the battery.
Dashboard-Percent does not tell you the amount of energy in the battery, it tells you how close to your "current full limit" it is, but it doesn't tell you what that "current full limit" amount actually is, and "full" *changes* meaning over time.

Thus, by definition, Dashboard-Miles is more meaningful than Dashboard-Percent.

If you disagree with this, you are disagreeing with the dictionary.

If you want to express an opinion, not a fact, that you personally prefer percent over miles, that's an opinion and you have every right to prefer percent over miles, but Miles is more meaningful than Percent, because it conveys more information to you.