Detecting Electric Vehicle (EV) devices


I have two BMW I3 (along with 2 identical GE chargers) and after a few weeks Sense has picked them up (they are not recognized as distinct devices just yet, but as a single charger). However, the reported kW and kWh is very low (less than 10% of actual) and the stop times are off as well. The “Now” screen of course reports the correct power draw, but assigns only a small amount to the EV and rest to “Other”).

Hey @larry.jolliffe,

Sorry to hear that! I’ve let our data science team know. It’s possible that the detection algorithm for the i3 needs some refinement. It’s also possible that as Sense sees the device run more and gets more data, it will be begin to attribute the rest of the wattage.


Larry congratulations on your successful i3 charging identification! I’ve been waiting 1 year (no joy). May I ask how long it took to identify the vehicle(s)? The attached is my charging session at home using the BMW charger (10pm nightly when necessary). I’m trying to figure out if the driving variable is the car or the charger? Thanks.

Hello mstraka606-

In my case, I think Sense found the chargers/cars (as noted above…there are actually 2, but they are identical) in about 2 weeks. But, the identification is limited to noting when the car gets plugged in. The kW, kWh, charging complete, and total charging times are not in line with reality (right now its showing an average rate of 489W, when it should be much much higher) . But, its a start!

FYI, my chargers are GE Wattstations (but Im not sure how much the charger plays into the electrical signature?)

Just to clarify, does “thrown a wrench” mean in this case that no one is currently actively working on it?

davidferri David, I’ve had my Sense operating for about 9 weeks now, and charge my Tesla Model S nightly. Sense still hasn’t picked up my S either…

It just means that the team is exploring other ways to track the LEAF’s off and that it’s unfortunately taking longer than we’d like to develop a detect for the LEAF.

Sense has recently identified part of my 2017 Chevy Volt’s charging-related operation: battery temperature maintenance due to cold temperatures (the existence of which was unknown to me until Sense found it):

Curiously, it was identified as an unknown heating element (which at least gets the idea of what it’s doing). The trace shows the typical “stair-stepping” rise that an actual charge cycle exhibits. Plus a curious little dip on the way up. When the Volt is plugged in and fully charged, the cycle runs for 10m every hour or so (it’s been single digits here for some weeks). It’s used about 22kWh this month.

Looking forward to true Volt charge detection (any updates?), but it’s heartening to know Sense can recognize at least something coming over the charge cable.


I’ve had Sense installed for almost 3 months. My 2016 Model S 60D with HPWC is set to charge at 11 PM each night when the time of use rates are lowest. To my eye, this seems like an extremely obvious usage pattern. Any thoughts as to why Sense isn’t identifying this device?

1 Like

Hi David,
Do you have a 72A or a 80A (dual) charger in your S ? Sense picks up the 80A charge for our 2014 P85D S correctly, most of the time. It does not detect the 72A charge cycle for our 2016 P90D X. I can’t remember when Tesla made the change from 40/80A, to 72A internal chargers, but I think it was late 2015.

Don’t forget all new UMC’s are stuck at 32 amps, no more 40s.

Yes, SENSE needs to deal with multiple possible signatures! (old UMC, which I have=40A capable, new UMC=32A capable, TWC=80A max output). Model-S/X being charged could have either 1x32A, 1x72A, (2016+)1x40A or 2x40A (pre-2016) ac-dc internal chargers. And now TM3 with ??? for chargers.

I have a Chevy bolt and a ford focus electric. Hoping sense can learn them soon! I connect them both using emotorwerks evse.

1 Like

good to know its being worked on…

I have a chevy volt, 7 months, alas it also is not detected…

220v charger on a dedicated 20amp 2 pole breaker.
I use the volt “time to charge” feature to make the car ready for 4am each day, so it uses the discounted evening power, it starts its charge cycle based on how much time it will take for the battery to be full by 4am, so start times vary…

I’m so disgusted with the lack of a desktop app its 2 or 3 months between even turning on the app on the phone, just like the people who write the Sense App - I have large dual screen monitors and would like to see all the data, not just tiny bites of it while zooming and scrolling on a 5 inch phone screen

whine time over… get back to work… get the car chargers discovered, and make a desktop version of Sense app

Hi @vegasmassimo -

Apologies for the delay on the Chevy Volt. Please know that we’re actively working on these, and grateful for your patience in the meantime. One of the challenges is that many of these chargers don’t present strong transients at the beginning or ends of their charging cycles; they ramp up over a long period of time. Are you seeing similar behavior when you watch the charger on your Power Meter in the Sense app? We’re adjusting our detectors to adapt to this, so hopefully we will have a solution for you at some point.

To your note on the desktop app, have you seen the web app we released in December?

It’s only the first version, so we have many improvements planned across 2018 to make it even more useful, but we’d encourage you to check it out and send along your feedback. You can log in at

Hope that’s helpful!

1 Like

I’m another Volt owner who would love to see that charger classified. It’s my largest power user and I’d like to be able to track its consumption separately from the rest of the house.

@BradAtSense Yes, I notice that the volt charger ramps up at the beginning of its charge cycle over a ~10 second period, and also ramps down at the end of its charge cycle over a much longer period (15 minutes or so). See attached.


I have a Bolt EV and an ‘Electric Vehicle’ is showing up in the ‘up next’ section of Device Detection… Looks like more models are being discovered now

1 Like

I have the 48A standard onboard charger with the wall connector installed on a 60A circuit, so I’m able to charge at the full 48A rate. This matches the energy usage spikes starting at 11 PM each night in the graph above, which are topping out somewhere over 11 kW. (48A x 242V ~= 11.6kW)

I’ve now had the Sense installed for over 4 months, and still no detection of the Tesla. Very frustrating. Sense staff - are you listening? I’m happy to help you collect data to be able to properly identify this signature, which should be very common among Tesla owners. I removed my Open Energy Monitor system when I installed the Sense, but could put it back in to get isolated energy usage logs for the dedicated EV charging circuit & send those to you for analysis. That system has dedicated CTs for individual high-load circuits like the charger.

1 Like

We now have 3 different Teslas, an S (dual charger 80A), an X (72A) and a 3 (48A) and have two HPWC each on their own 100A line. Right now Sense is only able to identify the 80A charging cycle. But I’m guessing that the guys at Sense have to do a custom model for each flavor of charging - not possible to do a generic Tesla or EV charging model, even though the parameters of a charge cycle make them painfully obvious to the human eye looking at the Power Meter view.

Are there things that EV owners can do to make detection quicker? For example, does detection work faster if the max charger current is not changed or the time of start is always the say time of the day. Are EVs detected faster if they charge everyday vs a few days per week?

As noted by others, EVs use a big chunk of energy and getting EVs detected is a huge part of the energy picture. Also, EV owners are generally early adopters. Sense needs the passionate support of early adopters to move into the main stream.