Fluctuating predicted range and even SOC when turned off

[SanSerif]

I've now got ConnectedDrive up and running and can amuse myself by watching the car's status as it charges. I've also started recording a whole load of metrics at the start and end of journeys. And these combined observations are raising some questions/oddities.

For example, it'll charge up to 100% and predicted range of, 78 miles. But if I look later it will sometimes read 100% and 82 miles. And if I check first thing in the morning it'll say something silly such as 100%, predicted range 95 miles. Unlocking the car this promise continues - but once I've turned everything on, the range will be down to, say 85 miles.

I've also noticed that the range estimate seems flakey - at least for short (sub 5 mile) journeys. I often travel 5 miles - but "lose", say, 8 miles of predicted range. I have yet to do any properly recorded tests on longer trips (hopefully tonight - I have a 60 miler).

Finally, I've noticed "loss" of SOC while parked unplugged in at work. As much as 2.5% vanishing.

Early days yet and I'll keep recording and monitoring. Without precise recording I wasn't aware of these differences and strangenesses - as everything was *broadly* correct. I'm also recording external temp - so some may be explained by temperature changes throughout the day.

 

[TomMoloughney]

I've now got ConnectedDrive up and running and can amuse myself by watching the car's status as it charges. I've also started recording a whole load of metrics at the start and end of journeys. And these combined observations are raising some questions/oddities.

For example, it'll charge up to 100% and predicted range of, 78 miles. But if I look later it will sometimes read 100% and 82 miles. And if I check first thing in the morning it'll say something silly such as 100%, predicted range 95 miles. Unlocking the car this promise continues - but once I've turned everything on, the range will be down to, say 85 miles.

I've also noticed that the range estimate seems flakey - at least for short (sub 5 mile) journeys. I often travel 5 miles - but "lose", say, 8 miles of predicted range. I have yet to do any properly recorded tests on longer trips (hopefully tonight - I have a 60 miler).

Finally, I've noticed "loss" of SOC while parked unplugged in at work. As much as 2.5% vanishing.

Early days yet and I'll keep recording and monitoring. Without precise recording I wasn't aware of these differences and strangenesses - as everything was *broadly* correct. I'm also recording external temp - so some may be explained by temperature changes throughout the day.

Yes, I would expect this. Every EV I've driven acts in a similar fashion. The state of charge and predicted range are far from exact measurements and the car is trying to use whatever information it can including new readings to offer you the best "guess" it can.

This is one of the reasons I have lobbied for a proper State of charge gauge (which we won't have in the US). Not because I will depend entirely on it, but because it gives on one more thing to look at, besides the guess the car makes with it's predicted range. That may well be pretty accurate, and BMW has been pounding their chest broadcasting how much effort they put into making the range assistant better and more accurate than any other EV on the market, but I still want my SOC gauge as a backup, so I can use both and draw my own assumptions on whether or not I can make a planned trip.

 

[i3atl]

This is one of the reasons I have lobbied for a proper State of charge gauge (which we won't have in the US).

Tom, I was under the impression that BMW is reconsidering this, given the feedback that they've received (and based on the same process that Nissan went through with the lack of SoC in the LEAF). Are you of the opinion that the conversation is dead, and we won't be getting a way to see SoC?

 

[TomMoloughney]

This is one of the reasons I have lobbied for a proper State of charge gauge (which we won't have in the US).

Tom, I was under the impression that BMW is reconsidering this, given the feedback that they've received (and based on the same process that Nissan went through with the lack of SoC in the LEAF). Are you of the opinion that the conversation is dead, and we won't be getting a way to see SoC?

I don't believe it will be on the cars in the US at launch, but it isn't dead. If BMW gets enough feedback that we want it it can be added in a software update that your dealer will perform. They need to continue to hear that people want it though - so if you want it also, let them know!

Months after finding out that it wouldn't have a proper SOC display I am still amazed. For five years I've been driving BMW's test EV's and they have always had a SOC display, and the users have consistently given feedback that we like it and want it on future EV's. I am completely puzzled about this and very disappointed. Someone was asleep at the wheel.

 

[i3atl]

I don't believe it will be on the cars in the US at launch, but it isn't dead. If BMW gets enough feedback that we want it it can be added in a software update that your dealer will perform. They need to continue to hear that people want it though - so if you want it also, let them know!

Months after finding out that it wouldn't have a proper SOC display I am still amazed. For five years I've been driving BMW's test EV's and they have always had a SOC display, and the users have consistently given feedback that we like it and want it on future EV's. I am completely puzzled about this and very disappointed. Someone was asleep at the wheel.

What is the most effective way to provide feedback like this related to the i models?

I'm guessing the absence of the SoC display is due to BMWs desire to obscure the details behind how the car is operating, so that they can interpret the "raw" info for you and present it in the way they want. They've been doing this ever since I've owned their cars - for example, moving from temperature gauges that reflected the actual water temp, to buffered gauges that sat directly in the middle of the range when temps were within what they considered a normal range, to (now) deleting the temperature gauge altogether - so you only get a warning light when there's a problem.

If they don't change their mind about the SoC, I imagine many people will resort to coding to enable the display (if possible) or adding a display via data from the OBD port (assuming BMW exposes SoC data through OBD). It would be a shame to be forced to go this route.

 

[agj]

I don't believe it will be on the cars in the US at launch, but it isn't dead. If BMW gets enough feedback that we want it it can be added in a software update that your dealer will perform. They need to continue to hear that people want it though - so if you want it also, let them know!

Months after finding out that it wouldn't have a proper SOC display I am still amazed. For five years I've been driving BMW's test EV's and they have always had a SOC display, and the users have consistently given feedback that we like it and want it on future EV's. I am completely puzzled about this and very disappointed. Someone was asleep at the wheel.

I recognize that at the beginning I did not understand all the fuss about the absence of a SoC; as I explained, the i3 is my first EV ever and I could not miss what I never experimented before on a previous EV. Besides, I thought that the predictive range estimate given by the car was precise enough.

And it is definitely very precise but I realize that it can only work on the basis of what it already knows (past milage driving style, temp, charge, topography in case of a given destination programmed on the Nav).

But it can be very tricky to rely on that predictive range only on one of those days when even you, as a human, don't know what the next move will be ("kids and friend need an unplanned lift to the other end of the city with the trunk filled with two heavy tube amps", "mother-in-law just called to get the pie she just prepared before I can recharge", "today I feel like a want to drive like petrol head and I don't want to be overtaken by this black Audi", etc.). Though the actual range delivered by the battery has been able to cope with quite a few of those "one-of-these-days" with flying colors since I own the car, the predictive range display has been struggling hard to adapt to the unexpected driving style, weight, charging conditions, etc. and, sometimes, was even completely useless.

A more readable SoC, with percentages like in the iRemote app, would have been very handy.

So, you can count me in for the lobbying work here in Europe in order to get some sort of decent SoC back on the display

 

[TomMoloughney]

I don't believe it will be on the cars in the US at launch, but it isn't dead. If BMW gets enough feedback that we want it it can be added in a software update that your dealer will perform. They need to continue to hear that people want it though - so if you want it also, let them know!

Months after finding out that it wouldn't have a proper SOC display I am still amazed. For five years I've been driving BMW's test EV's and they have always had a SOC display, and the users have consistently given feedback that we like it and want it on future EV's. I am completely puzzled about this and very disappointed. Someone was asleep at the wheel.

I recognize that at the beginning I did not understand all the fuss about the absence of a SoC; as I explained, the i3 is my first EV ever and I could not miss what I never experimented before on a previous EV. Besides, I thought that the predictive range estimate given by the car was precise enough.

And it is definitely very precise but I realize that it can only work on the basis of what it already knows (past milage driving style, temp, charge, topography in case of a given destination programmed on the Nav).

But it can be very tricky to rely on that predictive range only on one of those days when even you, as a human, don't know what the next move will be ("kids and friend need an unplanned lift to the other end of the city with the trunk filled with two heavy tube amps", "mother-in-law just called to get the pie she just prepared before I can recharge", "today I feel like a want to drive like petrol head and I don't want to be overtaken by this black Audi", etc.). Though the actual range delivered by the battery has been able to cope with quite a few of those "one-of-these-days" with flying colors since I own the car, the predictive range display has been struggling hard to adapt to the unexpected driving style, weight, charging conditions, etc. and, sometimes, was even completely useless.

A more readable SoC, with percentages like in the iRemote app, would have been very handy.

So, you can count me in for the lobbying work here in Europe in order to get some sort of decent SoC back on the display

Thank you. What you describe is perfectly understandable and I totally agree that many people who haven't had electric driving experience wouldn't think a proper SOC gauge is necessary because they could just use the estimated range gauge. But then you start driving your EV and realize it's not that simple and the range can greatly vary from day to day depending on all kinds of conditions. Having a SOC does indeed help you. Once you've had the car a while you learn how much SOC you'll use under different conditions and such. Having both the range estimator and the SOC at your disposal is definitely better than only having only one of them. BMW made a mistake. Lets hope they correct it. If they don't People will definitely hack the car and do it themselves like i3atl suggested. It's a shame we would have to do that to a premium electric car when much less expensive EV's offer it standard.