TMC is an independent, primarily volunteer organization that relies on ad revenue to cover its operating costs. Please consider whitelisting TMC on your ad blocker or making a Paypal contribution here: paypal.me/SupportTMC

Evaluation and Observations with v6.1 Trip Energy Prediction

Discussion in 'Model S: User Interface' started by Vger, Jan 17, 2015.

  1. Vger

    Vger Active Member

    Joined:
    Apr 10, 2009
    Messages:
    1,683
    Location:
    Salt Spring Island, BC, Canada
    I though it would be useful to start a new thread specific to the Energy Prediction feature of v6.1. We just did our first road trip with the feature and I have this to say: EUREKA!! This is the capability that will banish range anxiety once and for all!

    Here are some initial findings and undocumented features; I invite others with the software to post additions, corrections, etc. and those still without to propose experiments for further study. As a start:


    1. The prediction seems to be based on realtime measurements of total energy consumption rate (both propulsion and non-propulsion loads), as well as a high resolution GIS elevation profile of the route, and known properties of the drivetrain (lookup table for different models?).
    2. The prediction seems to assume travel at the same speeds (~legal limit) used to predict ETA.
    3. So far, there seems to be no evidence that it take into account weather conditions, especially wind. In fact, our experiences yesterday indicate that the prediction will be overoptimistic in head winds and pessimistic in tailwind conditions.
    4. The energy use graph has green, yellow and red regions which correspond to the colours of the battery meter on the instrument panels. The boundary between green and yellow is at about 20%, and the boundary between yellow and red is at exactly 7% SOC.
    5. If you are predicted to arrive at your next charge stop with <1% charge remaining, you will get the RED warning "Charging needed to reach destination."
    6. If you are predicted to arrive at your destination with <8% charge remaining, you with get the YELLOW warning "Drive slowly to reach destination."
    7. If you are predicted to arrive with 8+% remaining energy, you are considered "good to go" at that time and get no warnings.
    8. The entire prediction is recalculated quite often, at least every 15-30 seconds, so changes in speed, climate control settings, etc. are reflected very quickly, giving you almost immediate feedback.
    9. THIS IS A BIG DEAL: The prediction is re-calculated even during charging! So at a supercharger stop, if you put in your next charging destination, you will get a direct indication of when you have charged sufficiently to reach it— no mental math required! [The crowd roars!]
    10. Our tentative observation is that the prediction's calibration for elevation changes is very slightly off. It is slightly optimistic on uphills (underestimating energy draw), and slightly pessimistic on downhills (underestimating region impact). This is erring on the side of safety, so no too bad.
    11. The software does not understand ferry travel at all. It assumes that you are driving over the water, and using at least as much energy as you would driving. So any prediction including a ferry route is useless (nota bene B.C., WA, Scandinavians, etc.).
    12. We used the (+) zoom mode all the time. This allows much closer inspection of the difference between original prediction (grey line) and actual progress (coloured line).
    13. If you cancel the route and then reenter it, you will lose the part of the energy graph from your original start point to your present location. This can be used to advantage to zoom up on a more critical section at the end of your trip, especially if things are getting dicey.

    Here are a couple illustrations of what happens during (super)charging:

    First image is near start of charging session, and second is near the end. This is for the 170 mile run from the Ellensburg, WA to Burlington, WA superchargers.
    IMG_0030.jpg IMG_0033.jpg
     
  2. jerry33

    jerry33 S85 - VIN:P05130 - 3/2/13

    Joined:
    Mar 8, 2012
    Messages:
    12,753
    Location:
    Texas
    This isn't surprising because wind and weather often vary with microclimate. Unless there was a mini-weather station on the car it would probably do more harm than good.
     
  3. Lump

    Lump Active Member

    Joined:
    Mar 31, 2013
    Messages:
    2,102
    Location:
    So. Cal.
    I don't have 6.1 yet so this helps a lot, hopefully our smartphone app will provide the same charging session graph soon.
     
  4. apacheguy

    apacheguy Sig 255, VIN 320

    Joined:
    Oct 21, 2012
    Messages:
    3,709
    Location:
    So Cal
    Yes, that is impressive, but it don't like the underlined part. Reason is here in CA we often travel at 10 MPH over the speed limit. If it's basing it's calc off that, then it will be horribly inaccurate in telling me how much I need to get to the next SpC. So mental math will still be required. I suggest that it use average efficiency numbers as a more accurate unit.

    Agreed. But it could do temperature quite easily.

    Wow, this is a loaded thread. That's a fantastic idea. That way I don't have to keep running out to the car to check the trip calculation.
     
  5. VirtualE

    VirtualE Member

    Joined:
    Feb 10, 2013
    Messages:
    50
    Location:
    Houston, TX
    This is a great summary and report.
    While I never head range anxiety, due to head planning. This feature will help a lot. Specially because it will allow me to know if I need to stop to charge or if I can stretch the trip.
    in terms of micro climate, I am sure that in the future Tesla could add this feature and variables. The information exists. There are already apps that give you precise wind direction and speed, and of course temperature.
     
  6. Todd Burch

    Todd Burch Electron Pilot

    Joined:
    Nov 3, 2009
    Messages:
    4,733
    Location:
    Smithfield, VA
    Cool--looking forward to receiving 6.1.

    Instead of using speed limit, I'd like to be able to set an offset from the speed limit (like speed assist currently uses) for all ETA/energy requirement predictions. I think that would be more useful to people.
     
  7. ecarfan

    ecarfan Well-Known Member

    Joined:
    Sep 21, 2013
    Messages:
    10,338
    Location:
    San Mateo, CA
    I used the new Energy display for the first time today during a 140 mile roundtrip that included a 1,000 ft elevation gain. Climate was mild with no wind and temps in the 50's (F). The new display was very accurate with the prediction closely matching actual usage.
     
  8. apacheguy

    apacheguy Sig 255, VIN 320

    Joined:
    Oct 21, 2012
    Messages:
    3,709
    Location:
    So Cal
    Indeed. I was thinking about it and realized that if you could input your average speed that would yield closer to true world predictions.
     
  9. ZBB

    ZBB Emperor

    Joined:
    Feb 27, 2013
    Messages:
    1,417
    Location:
    Scottsdale
    Thanks for the details Vger! Can't wait to get 6.1 and try those out.

    As as for the ferry issue... Nav systems never get that right anyway, so not that surprised that it thinks it will need energy to move over water. When I did the planning for our SSI visit last summer, I manually adjusted the EV Trip Planner estimates to take out the ferries -- no need to spend extra time at a supercharger for energy not needed...
     
  10. Saghost

    Saghost Active Member

    Joined:
    Oct 9, 2013
    Messages:
    3,003
    Location:
    Delaware
    It's a great start, without a doubt, and something I haven't seen on other EVs. As the OP said, this has the potential to be the real answer to range anxiety.

    Pulling the weather every twenty or fifty miles along the route would probably give a pretty good picture, I would think - allowing the car to predict not only the effect of wind, but also the impact of HVAC.

    As for speed effects, I'm thinking it should base the driving speed on the offset chosen for the speed limit assist for newer cars (maybe create a separate offset option for older cars?) - that should give an individualized result that matches the driver's habits.

    Updating in real time while charging is a great addition - part of the way to what I wanted for the 'road trip' update.

    Now all you need is a time estimate after you plug in (once you plug in it should be quite predictable, since it'll know if you're a second car or there are equipment limits,) and something for the mobile app - ideally a countdown clock but minimally a notification sent to the phone when you hit a selectable predicted energy at destination.
    Walter
     
  11. Ven Rala

    Ven Rala Member

    Joined:
    Nov 17, 2012
    Messages:
    378
    Location:
    Montgomery County, Md
    I have started using it on my daily commute and it is a great feature. I just wish I had it 3 weeks ago on my 2600 mile round trip to Florida
     
  12. jerry33

    jerry33 S85 - VIN:P05130 - 3/2/13

    Joined:
    Mar 8, 2012
    Messages:
    12,753
    Location:
    Texas
    Agreed it's a great start, but I've gone on a number of trips and range anxiety has been far less than in a gas car. In a gas car you have range anxiety before every fill. In an electric car you preplan your stops and no range anxiety occurs.
     
  13. BG121

    BG121 Member

    Joined:
    Jul 7, 2013
    Messages:
    36
    Location:
    Redwood City CA
    The next iteration should allow you to select your average trip speed in relation to the speed limit (+0, +5, +10, or +15 mph over the limit) so the trip graph can more accurately predict energy needed to reach. This would be helpful for areas where you need to drive a bit faster than the speed limit to keep up with traffic.
     
  14. mackgoo

    mackgoo Member

    Joined:
    Oct 18, 2014
    Messages:
    134
    Location:
    SoCal
    Once on the road I'm pretty sure it takes your speed into consideration and how that affects you at the end. The NAV did that for time remaining.
     
  15. BG121

    BG121 Member

    Joined:
    Jul 7, 2013
    Messages:
    36
    Location:
    Redwood City CA
    This is true but the point I was trying to make is that the trip graph because even more valuable in planning SoC ahead of time if you can indicate expected speed during the trip
     
  16. Saghost

    Saghost Active Member

    Joined:
    Oct 9, 2013
    Messages:
    3,003
    Location:
    Delaware
    As I pointed out upthread, for new cars there's an obvious place to grab this already - the offset you choose for the speed limit assist. For older cars, it might be helpful to add it explicitly for this function.
     
  17. eye.surgeon

    eye.surgeon Member

    Joined:
    Nov 18, 2014
    Messages:
    707
    Location:
    California
    There's a difference between range anxiety and inconvenience of refueling an ICE car. You can preplan stops in an ICE car too, if you want. Let's keep it real, you have to be extremely absent-minded to have range anxiety in an ICE.
     
  18. gpetti

    gpetti Active Member

    Joined:
    Nov 6, 2013
    Messages:
    1,573
    Location:
    Calgary, Canada
    thanks for this excellent summary. Collective learning is a great thing. When I'm finally able to use it I'll be ahead of the curve.
     
  19. David99

    David99 Active Member

    Joined:
    Jan 31, 2014
    Messages:
    2,051
    Location:
    Brea, Orange County
    Since I don't have 6.1 yet, can you now enter a route with multiple waypoints? This is one feature that's missing and often needed for road trips.
     
  20. ecarfan

    ecarfan Well-Known Member

    Joined:
    Sep 21, 2013
    Messages:
    10,338
    Location:
    San Mateo, CA
    I fortunately, no.
     

Share This Page