+10% Energy Efficiency after 2.24.102 Update

[ZeroDarkSilver]

My commute is 75 miles, of which 95% is driven on Autopilot. Prior to 2.24.102, it was consistently taking 330 Wh/mi, but ever since that update, the energy usage of the trip has dropped to 300 Wh/mi.

Since 2.24.102, I've felt that Autopilot accelerates more smoothly. For example, when traffic clears ahead, the car now ramps up to its maximum set speed more gradually than before. And I think that braking, when traffic-ahead slows, might also be more incremental and less last-minute, but I'm less sure of this effect.

I don't have hard data on this, but Autopilot seems like it would amenable to optimization to improve energy efficiency, and would be something worthwhile for Tesla work on. Maybe they are.

Anyone else notice a change?

 

[Topher]

Since 2.24.102, I've felt that Autopilot accelerates more smoothly. For example when traffic clears ahead, it ramps up to its maximum set speed more gradually than before. I think the braking, when traffic-ahead slows, might also be more incremental and less last-minute, but I'm less sure of this effect. I don't have hard data on this, but I think it's certainly possible that Autopilot optimizations can improve energy efficiency.

Hard to imagine that even that admittedly better algorithm for accelerating and braking could make a 10% difference. Good to hear though.

Thank you kindly.

 

[chillaban]

Funny you should mention this. I just got 2.24.102 today as well, and arrived home and noticed I used 2% less battery capacity than before which is around 10%. Autopilot was quite smooth dealing with cars cutting in, too.

 

[MDMGSO47]

My commute is 75 miles, of which 95% is driven on Autopilot. Prior to 2.24.102, it was consistently taking 330 Wh/mi, but ever since that update, the energy usage of the trip has dropped to 300 Wh/mi.

Since 2.24.102, I've felt that Autopilot accelerates more smoothly. For example, when traffic clears ahead, the car now ramps up to its maximum set speed more gradually than before. And I think that braking, when traffic-ahead slows, might also be more incremental and less last-minute, but I'm less sure of this effect.

I don't have hard data on this, but Autopilot seems like it would amenable to optimization to improve energy efficiency, and would be something worthwhile for Tesla work on. Maybe they are.

Anyone else notice a change?

Do you have DRLs?

 

[ZeroDarkSilver]

Do you have DRLs?

Yes, DRLs lit up for me after 2.24.30, but then shortly thereafter one of them partly failed (DRL: Now you have it, now you don't...). Why do you ask?

 

[msnow]

Yes, DRLs lit up for me after 2.24.30, but then shortly thereafter one of them partly failed (DRL: Now you have it, now you don't...). Why do you ask?

Dumb question: Doesn't everyone have DRL's?

 

[ZeroDarkSilver]

Dumb question: Doesn't everyone have DRL's?

The 2016 refresh model with LED headlights didn't have working DRL until the 2.24.30 software update.

 

[chillaban]

The 2016 refresh model with LED headlights didn't have working DRL until the 2.24.30 software update.

Dont' forget... Tesla reminds you that they're called the "Signature Lights", not DRLs, with the refreshed Model S

 

[xG35]

I thought the car was smoother and faster.. just got it this afternoon.. so not sure if it is more efficient.

 

[Patrick W]

Can't say. I'm still back at 2.20.30. No idea why some receive updates whiles others do not.

 

[msnow]

Dont' forget... Tesla reminds you that they're called the "Signature Lights", not DRLs, with the refreshed Model S

Interesting, I thought I saw the term used in both the manual and controls and settings menu. I'll have to look again.

 

[RossS]

Pretty sure this is a result of the Wh/mi calculation being tweaked, rather than any real energy efficiency improvement. I recall reading something on another thread about the Wh/mi not calculating properly on newer Model S vehicles with software limited batters (for example, I have an updated facia S70 that actually has a 75 KWh battery).

That said, I'm definitely noticing seemingly improved efficiency since installing this update (as described in the OP). Previously, I was wondering why no matter how much I tried to drive consciously, I couldn't consistently get my Wh/mi under 300!

 

[ZeroDarkSilver]

Pretty sure this is a result of the Wh/mi calculation being tweaked, rather than any real energy efficiency improvement. I recall reading something on another thread about the Wh/mi not calculating properly on newer Model S vehicles with software limited batters (for example, I have an updated facia S70 that actually has a 75 KWh battery).

That said, I'm definitely noticing seemingly improved efficiency since installing this update (as described in the OP). Previously, I was wondering why no matter how much I tried to drive consciously, I couldn't consistently get my Wh/mi under 300!

If by Wh/mi calculation you literally mean the simple division of the Wh count by the mi odometer, that hasn't changed and has been correct the whole time. Also the odometer (I'm referring to the trip computer here), seems to be accurate.

The thing that seems to have changed is the kWh count tracked by the trip computer, for what are basically the same trips before and after. It's possible this has changed (I don't have a software-limited battery), but the reported kWh counts seem to match up with actual draw from the charger so far, so the only explanation is that efficiency went up.

My theory here is that lead-footedness can make a big difference in energy consumption per mile, and they are improving Autopilot's lead-footedness over time. That said, I think speed/wind resistance makes the biggest difference so it's possible there's just been more traffic recently and I'm going at a lower overall speed.

After I get my next electricity bill I'll be able to tell more definitively.

 

[chillaban]

I'll be able to tell in another week or so. I very consistently use 20RM on the way to work and 15RM on the way home. With a few more data points I'll be able to see if that changed at all, or if it's just the energy graph's computations that changed.

 

[chillaban]

So, first data point: It usually takes 20 rated miles for me to drive to work (15.8 real miles). Very consistently 20. But today, I was able to do so in 18 rated miles.

It's not conclusive with a single data point, but it does roughly line up with a 10% efficiency difference. We shall see how that holds up over time. But definitely, the Wh/mi computation was similarly 10% lower.

 

[msnow]

So, first data point: It usually takes 20 rated miles for me to drive to work (15.8 real miles). Very consistently 20. But today, I was able to do so in 18 rated miles.

It's not conclusive with a single data point, but it does roughly line up with a 10% efficiency difference. We shall see how that holds up over time. But definitely, the Wh/mi computation was similarly 10% lower.

Why don't we look in the release notes to see if they made improvements to this algorithm? Oh wait, we don't get release notes.

 

[Mattzilla]

I had 2.20.30 until last Friday, when 2.24.102 came through for me. My DLR did not work before the update while the lights were set to auto (I had to manually engage parking lights each time I drove the car to get DLR to work). Now they work in auto, but only when range mode is disengaged.

I haven't noticed any efficiency improvement. In fact, my efficiency seems to be worse since the update. I was averaging in the high 100s or low 200s before. Now I get around 250-280 wh/mile.

 

[chillaban]

Why don't we look in the release notes to see if they made improvements to this algorithm? Oh wait, we don't get release notes.

Haha. To rehash a point from that thread, I believe Tesla would get into a lot of EPA trouble if they attempted to claim higher economy via a software update ;-)

 

[chillaban]

I haven't noticed any efficiency improvement. In fact, my efficiency seems to be worse since the update. I was averaging in the high 100s or low 200s before. Now I get around 250-280 wh/mile.

Whoa! How did you get high 100's or low 200's? I've not even seen that cruise-controlling on a flat road without traffic at 60mph!

 

[Alketi]

Can't say. I'm still back at 2.20.30. No idea why some receive updates whiles others do not.

It's very common for software companies to stage OTA updates, so as to not take down the entire population of users should there be a problem. The initial rollout may be to 1% or 10% of users, with a few days until the next stage, to see if there are any issues reported.