LEAF Guess o Meter GOM comparison to Tesla

[sp4rk]

Having "experienced" the Leaf's "Guess-O-Meter" ... for a few weeks.

I am very interested in how Tesla is doing this?

We'll not know I assume before test drives become commonplace or deliveries?

Anyone gotten any early ideas or knowledge yet?

 

[bonnie]

So assuming you mean range indicator (I haven't seen the guess-o-meter term before), I've found the Roadster range to be pretty accurate. I have both ideal range information (if I were driving conservatively) and estimated range information (based on how I'm currently driving). I would assume the same for Model S and Model X.

 

[stopcrazypp]

Having two numbers like the Roadster does, is probably the best approach. The ideal number only varies with capacity and not with the road conditions. So it's a reliable ideal number.

You would have to ask existing Roadster owners if the other estimated number is accurate.

 

[dhrivnak]

I have found the Roadster to be very accurate. Now in normal mode, not ideal, it can vary a lot as it is based on power used over the last 30 miles and that can vary a lot when you start off with a full battery. But then it settles down and is quite accurate by the end.

 

[Sparrow]

I find my Leaf's range estimate to be very disconcerting. In comparison our Roadster set to ideal miles is very accurate and even gives you something to shoot for while you are driving. As others have mentioned even the estimated range accuracy in the Roadtser is much better than the Leaf's.

 

[jerry33]

If the Roadsters range indicator was inaccurate, you'd see all kinds of posts on the forums just like you do on the Leaf forums. My assumption is that it's accurate enough to not cause complaints.

 

[ChadS]

I think if the Roadster's estimated miles display was all we had, there would be complaints. It definitely takes a long time to adjust if you change how/where you're driving; although it does seem less variable than the Leaf's.

But combine the Roadster's estimate with the ideal miles display (which is what I really use almost all the time), and it's just fine.

The Leaf's guess-o-meter, combined with a very coarse SOC display, causes range anxiety. I just had lunch with a friend that recently bought a Prius--he'd tried the Leaf (and it would have been perfect for him!), but was scared off by the wildly fluctuating GOM. I hear that sort of story all the time. The only "scary" incident my father has had in his Leaf was because of the GOM, even though I'd warned him not to trust it.

 

[TEG]

The range estimate gauges tend to have a proprietary algorithm that takes into account various things including battery pack voltage, recent driving energy usage, speed, and such.
Whatever they calculate, they tend to "smooth it" (average it) and pick an update frequency so that the gauge doesn't bounce up and down too frequently.
I think Nissan has acknowledged that they could do better, and plan to have improved readings in future models.
Tesla seems to have done a good job even with their first effort.
None of the manufacturers seem to be in a big hurry to explain exactly how they calculate this.

3rd party efforts like upcoming LEAFscan will help LEAF drivers get more information even if they can't get it from Nissan.

 

[vfx]

Leafscan reminds me that many Rav4 owners had a palm pilot mounted with full readouts as well.

When will EV makers realize that a certain segment want to see everything you can possible know about the cars energy use and generation. Whatever they include, someone will want something different. Just include it somewhere in the readouts. The early adopters of these cars are techies you know.

 

[VolkerP]

So assuming you mean range indicator (I haven't seen the guess-o-meter term before).

I'm late to this discussion but want to point to one of Tom Saxton's blog entries for anyone with questions what the Nissan LEAF has as a range indicator and what owners do about it.

 

[Talkredius]

So assuming you mean range indicator (I haven't seen the guess-o-meter term before), I've found the Roadster range to be pretty accurate. I have both ideal range information (if I were driving conservatively) and estimated range information (based on how I'm currently driving). I would assume the same for Model S and Model X.

The Ideal Range is one of the coolest and most needed thing for me as a long range driver. I always use my GPS to measure the distance between my next destination ((e.g. hotel or charging station) If the Ideal Range drops below the actual distance I know, I have to do something about my driving style, e.g start driving behind a truck get slower etc. Estimated Range is just a kind of motivation meter., I can see if I'm more efficient than an average driver.. After driving downhill a big mountain you might get an ER of > 500 km.. But when driving over some big mountains I do some math before to check if I'll arrive on the top of the mountain.

 

[SByer]

Actually, I think it's:

while (((EstMiles() / IdealMiles()) * IdealMiles()) < MilesToDest()) { SlowTheFDown(); }

But yeah, knowing ideal miles and how your estimated miles relate to it under the various driving conditions you've experimented with are critical for knowing how many miles you really, truly have left. And given the time length of the feedback loop on all of those pieces of information, the more detail, the better. The human brain can process large amounts of complexity and subtlety if consistency is also maintained.

 

[daniel]

Some early Prius (not mine, oddly) had a very inaccurate gas gauge, which people started calling the guess gauge. So the term guess-o-meter sounds good for the Leaf. They really should have had an SoC meter reading in percentage and one-percent increments. In effect, that's what the ideal range readout in the Roadster provides: a very fine indication of the state of charge, uninfluenced by driving style or road conditions. The estimated range, based on how you've been driving is a better indication of how far you're going to be able to drive, but with the caveat that changing conditions make it more of an estimate than a promise. I really like having both.

 

[rabar10]

Actually, I think it's:

while (((EstMiles() / IdealMiles()) * IdealMiles()) < MilesToDest()) { SlowTheFDown(); }

Your compiler will optimize the IdealMiles() reference out of your equation, methinks :wink:

 

[TEG]

They really should have had an SoC meter reading in percentage and one-percent increments...

Apparently it isn't so easy to keep track of exactly how much energy is stored in the battery except when it is left on charge for a while.
So, while driving with various power going in and out of the battery, they are constantly recalculating trying to do the best they can to figure out how much range might be left. But even Nissan engineers say that the battery pack sometimes surprises them with unexpected changes in SOC. I guess things like changes in temperature and humidity have an effect on range, and they can't readily predict exactly what will happen next as you drive around. The third party SOC meters people are making for the LEAF are helpful if you know how to interpret the data, but apparently Nissan felt the "raw data" wasn't too helpful to the average driver.

 

[ChadS]

[...] apparently Nissan felt the "raw data" wasn't too helpful to the average driver.

Understood that the SOC is hard to measure. Further understood that new EV drivers are not familiar with SOC (although they pick it up really quick; it's just like a gas gauge).

But Nissan decided to hide that "confusing" SOC, and instead took the SOC level, multiplied it by some mysterious measure of past consumption (for how long? weighted how? how will it compare to future driving?) and display that instead as the GOM. An uncertain number modified by another, even more uncertain number is not more helpful.

I like that they have the GOM; it has some use for new drivers, and for experienced drivers in new situations. But they need to display SOC as well--it is more important! That was the very first piece of advice Plug In America gave Nissan when we drove the cars; SOC is what we all use. (They actually do display it via the charge bars--but in a too-granular fashion, and it doesn't seem as prominent. Most new drivers I've talked to don't even realize the bars are there; they just look at the big, central GOM and are frightened by its fluctuations).

Nissan built a great car, but not having a good SOC meter is one of their few glaring mistakes. If they are worried about the units, fine, multiply them out to ideal miles like Tesla does. You can get used to different units; but you need a reliable measurement.

 

[TEG]

Agreed. I would trade the GOM value for the raw SOC values (even if they are a little squishy) any day.
In a way, it sounded like a rationalization... "We don't want to display it because sometimes it does odd things."
We end up with a wildly fluctuating "guess" value instead of a the slightly imprecise value.
Heck, fuel gauges in ICE cars are often imprecise. Many fuel gauges will swing when you go up and down hills and the fuel sloshes around in the tank...
Anyways, Nissan definitely got the message that customers would like for them to do better. I expect things to improve in future generations of their EVs.

 

[TonyWilliams]

Apparently it isn't so easy to keep track of exactly how much energy is stored in the battery except when it is left on charge for a while.
So, while driving with various power going in and out of the battery, they are constantly recalculating trying to do the best they can to figure out how much range might be left. But even Nissan engineers say that the battery pack sometimes surprises them with unexpected changes in SOC. I guess things like changes in temperature and humidity have an effect on range, and they can't readily predict exactly what will happen next as you drive around. The third party SOC meters people are making for the LEAF are helpful if you know how to interpret the data, but apparently Nissan felt the "raw data" wasn't too helpful to the average driver.

The 2013 LEAF will have a % readout in the dash, in addition to the GOM and 12 bar energy gauge. Presumably, the % readout will be exactly what we are already using with "Gidmeters".

With the percentage, you can get a great estimate of ideal range by simply multiplying that by 21kWh for a new battery at 70F degrees or higher, and then by the predicted economy.

So, a 50% readout would equal 11.5kWh (0.50 * 21) multiplied by 4 miles/kWh (typical for 65mph on level dry roads without heater use) would equal 42 miles of ideal range (4 * (0.50 * 21)).

That is mostly what the LEAF Range Chart does.

 

[Dave EV]

The 2013 LEAF will have a % readout in the dash, in addition to the GOM and 12 bar energy gauge. Presumably, the % readout will be exactly what we are already using with "Gidmeters".

And it still won't be quite as useful as having a gauge that reads "ideal miles remaining" since that gauge will also give you an indicator of battery capacity instead of relative battery capacity... I hope I'm wrong.

 

[TonyWilliams]

And it still won't be quite as useful as having a gauge that reads "ideal miles remaining" since that gauge will also give you an indicator of battery capacity instead of relative battery capacity... I hope I'm wrong.

I made a mistake; there is no way they will display Gids since when the 2013 LEAF battery tanks just like a 2011-2012, a full charge won't show 100%. So, it will be strictly SOC%.