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

Navigation-Based Projected Range Enhancement

Discussion in 'Model S: User Interface' started by ken830, Oct 9, 2012.

  1. ken830

    ken830 Model S (Res#P12,447)

    Joined:
    Jun 19, 2012
    Messages:
    858
    Location:
    San Carlos, CA
    First off, I'm not sure if this belongs here in the UI subforum or the Battery & Charging subforum. Mods, please move if I picked the wrong one.

    I have an idea for an enhancement. From one of the recent firmware update, Tesla is now providing two range estimates: Projected Range and Rated Range. Rated Range using the EPA's 5-cycle test to calculate range. Projected Range takes data from your last 30 miles of driving to predict the range.

    My idea would affect only the Projected Range estimate and only when the the navigation system is currently providing navigation/route guidance. The Projected Range prediction would take into account current traffic conditions and terrain information along the predicted route to further refine the projected range.

    Not only would this be helpful day-to-day to get more accurate predictions, but would be super-useful in resolving any range-anxiety when traversing a mountain pass. When climbing up the hill, the current system would paint a very grave picture of remaining range when, in reality, the downhill portion of the drive would partially mitigate that.

    One more feature that would come "free" with this is that the car can automatically tell you if you will be able to make it to your destination with your remaining charge.
     
  2. VolkerP

    VolkerP EU Model S P-37

    Joined:
    Jul 6, 2011
    Messages:
    2,455
    Location:
    Germany
    I agree that such an enhancement would be far better than extrapolating 1200Wh/mile going uphill and -800Wh/mile going downhill, which results in ridiculous values for projected range (Biggest and Smallest Wh/mi and Biggest Projected Range).

    One thing to add for the "mountain pass" use case: The software must check if you can reach the top. No regen will happen if you stop short of the pass!

    Actually the Project Better Place cars run an on board software that includes elevation changes into projected range and plans a trip accordingly. It suggests battery swap stations en route, in case of a Tesla vehicle make that superchargers.

    Why not take it a bit further? Have the range prediction software work with the navigation to calculate
    - where and how long to charge
    - recommend a speed / energy consumption rate for every leg of the journey
    - check back with traffic data if projected speed can be maintained
    - contact the superchargers in advance and reserve a charging time slot based on ETA

    You then can review the proposed route plan and adjust for longer stops and/or different locations. Example: get your lunch at your favorite restaurant in that other city, turning a half hour stop into a full hour one. The software will reschedule charge slots appropriately, as you go.
     
  3. ken830

    ken830 Model S (Res#P12,447)

    Joined:
    Jun 19, 2012
    Messages:
    858
    Location:
    San Carlos, CA
    I like your enhancements to my enhancements!

    In fact, I hope that the critical "car/safety" portion of the computer/software is well separated and protected from the infotainment system. That would allow Tesla to potentially open-up part of the system to 3rd-party software add-ons and plugins. There is a huge potential here. And there are many creative geniuses out there. Tesla can't hire them all.
     
  4. pilotSteve

    pilotSteve Member

    Joined:
    Jul 14, 2012
    Messages:
    579
    Location:
    Vancouver WA
    I like it! Google maps has topi data (at least on a 'grid' basis) that could be a starting place to estimate the vertical energy component of the route.
     
  5. dadaleus

    dadaleus 4GETOIL P85#S70,FdrX,S85D

    Joined:
    Aug 27, 2009
    Messages:
    884
    Location:
    Solana Beach, California, United States
    Some of this sounds much easier than other parts of it. But I love the direction. I'd think that incorporating topology and factoring into projected range or showing a projected use of rated range shouldn't be enormously difficult. If it knew speed limits (Tomtom and Garmin do, but not sure if there is a public DB available), you could adjust an assumption about how fast you are going relative to the limit (e.g. at the limit, limit +5%, limit +10% etc...). Incorporating traffic is probably a lot harder unless you use existing software (TomTom can dynamically route and adjust ETA on traffic, and it's pretty darn accurate).
     
  6. ken830

    ken830 Model S (Res#P12,447)

    Joined:
    Jun 19, 2012
    Messages:
    858
    Location:
    San Carlos, CA
    #6 ken830, Oct 9, 2012
    Last edited: Oct 9, 2012
    Incorporating traffic and speed limits shouldn't be hard.

    Google Maps and Navigation already knows typical speed limits and typical traffic data. That is how they always give you driving times with and without traffic.

    If you've ever used Google maps on a desktop, there is a little-known feature of traffic that lets you select a day and time to see typical traffic data at any point in time for any day of the week. This would be useful for road trip planning and range predictions. Imagine using projected traffic conditions to further refine range projections and ETAs.

    I do agree that they should implement the simple solutions first. Just using topographical data to give more realistic range predictions would be a huge leap forward!
     
  7. XrstalLens

    XrstalLens Model S P1327 VIN P01867

    Joined:
    Dec 30, 2011
    Messages:
    115
    Location:
    Kirkland, WA
    #7 XrstalLens, Oct 9, 2012
    Last edited: Oct 10, 2012
    Way back in October 2011 at the Factory Event my wife and I discussed this with one of the Tesla engineers. He specifically mentioned that they were looking at using topographical data in range prediction when you have your route entered into navigation. Frankly, I'm surprised it's not already available, but I have to assume that it's in the works.
     
  8. Cottonwood

    Cottonwood Roadster#433, Model S#S37

    Joined:
    Feb 27, 2009
    Messages:
    5,062
    Location:
    Colorado
    This is a great idea.

    It is pretty easy to get elevation (h) from the GPS. Potential Energy (PE) is just m*g*h. Taking an average load should almost always be +/-10% (~+/-500lb) of reality for m. Maybe the air suspension can do even do a better job. g is very close to a constant over the earth. Now you can remove very accurately the PE part of the energy use from the 30 mile trailing data. As has been said, getting elevation from the planned route should be pretty straight forward for predicting future PE needs. In fact, even without a planned route, I would rather have the predicted range displayed with trailing PE removed, and just assume flat ground ahead. That is better than around the world predicted range.

    There are two onboard nav systems: the one driving the instrument cluster route display and Google Maps on the 17" center console. The instrument cluster system has local data (works when there is no cell coverage and Google Maps paints and blank screen. The instrument cluster system also seems to have a pretty good, detailed elevation data base because it paints a good 3D synthetic view of the upcoming terrain. Hopefully that elevation data can be accessed easily.

    The "will I make it over the pass?" test is very important. On my drive from Boulder to Pagosa Springs, CO, last week, had I not charged before Wolf Creek Pass, the rated miles would have been -4 at the top of the pass, but positive as I entered town...

    One more note, the S does the same thing that R does on displaying rated miles on long downhills. Both won't show increased rated miles in the battery until the total added miles hit 3. They then add 3 to the rated mile total, and start over until you get another 3 added. Pretty weird, but I guess that they don't want the rated miles display to be too busy.
     
  9. kalikgod

    kalikgod Member

    Joined:
    Oct 8, 2012
    Messages:
    82
    Location:
    The Woodlands, TX
    I have actually been thinking about this also and have a few features to add:


    1) Use outside temp to estimate HVAC usage
    2) Use outside temp to adjust drag for air density changes
    3) Store a driving behavior factor for each driver and apply that to the range estimate
    4) Apply traffic congestion to average vehicle speed but also lower the speed for vehicle wakes (range is better with lots of cars doing 70, than doing 70 alone on a road)
    5) Check weather for wind speed and apply it (headwind, crosswind, tailwind)
    6) Check for rain and increase tire drag accordingly


    I would write an app for it myself, but it would be better done as an integrated system in the Nav.
     
  10. stevezzzz

    stevezzzz R;SigS;P85D;SigX

    Joined:
    Nov 13, 2009
    Messages:
    6,062
    Location:
    Colorado
    This is an application crying out for integration with the S software. Flight planning software on inexpensive portable GPS devices allows you to do all this and more, integrating winds aloft forecasts and climb/descent fuel burns in addition to cruise burn. The end result is a nav log with waypoints, ETA for each, fuel burns for each, etc. As has already been noted, this would go a long way toward taking the anxiety out of planning a cross-country trip in the S.

    On a related subject, this is pretty cool: Wind Map
     

Share This Page