Enough MS owners are chalking up enough miles now that it just might be time for a spread sheet to track battery degradation. The down side is the tendency of lazy "journalists" to cherry pick forum information and some even to use the information to paint unflattering and false pictures of Tesla and MS. The up side is that we owners can start to piece together a picture of all (or at least the ones that participate ) the forum member's experience with loss of range. This should include those that put a bunch of miles on their cars and those that do not along with those that top off all the time versus those that run the battery down a good bit before charging. I would like this information. Can we do this in a copyrighted fashion such that the information is not used improperly?
Nope (not by your definition of "improperly"). But we should do it anyway. We have nothing to hide, and we need it.
My question is how do you measure it? The computer generated estimate can change as we saw a drastic (and needed) change from when the 'rated' range was showing a drastically reduced and incorrect value. What we really need for data points is data on the actual battery capacity. At the very least, I would suggest people watch the 'ideal' range rating. Better yet, drive the car to empty and see how far you got. However, that is a pain with an EV, so perhaps to 5 mile range? I agree more information is better, and a way to measure the actual battery capacity would work best.
I was simply thinking starting by tracking reduction in standard and range charge available miles (?).
If you did, you would want to measure the things that impact range/capacity, so you could decipher what within "normal" and what would be an anomaly. Things like month/temp, miles driven/mo, avg. kWh/mile, etc.
If we are going to do this some rules and data points need to be adhered to: 1. Indicated range needs to be recorded immediately following completion of charge cycle. 2. Data should include both range and standard for both ideal and rated range settings. 3. Miles on vehicle 4. Average temperature during charging.
Which will result in bad data/trends if 'standard' or 'range' percentages change, or if the formula to calculate standard or range (less likely) changes, or if environmental factors that go into the formala changes. To do this right, you need the actual battery capacity. At the very least I would suggest only using 'ideal' range shown on a full/'range' charge. If there were a battery capacity reading made available to us that would work perfectly and I would encourage Tesla to make that available to owners.
This seems a bit extreme. I've only ever seen people quote rated range. I've never had my car set to show ideal range and never plan to, so I simply wouldn't have data to contribute with this requirement.
I've been logging data from my Model S via the REST API for a month or two, so I have a record of the charge state at the end of every charge. This information includes rated miles, ideal miles, and charge percentage. Additionally, I think I can log the odometer from the streaming API, whenever the car is driving, although I am not doing that right now. Unfortunately, there are no provisions for third-party access in the API, so if we wanted one person to collect data from multiple cars, we'd have to share our passwords. But we could create a script to allow people to easily collect their own data, in a standard format, and then we could compile this into a single spreadsheet.
I disagree, rated range is subject to change. It already has in the past. While some may view it as extreme, it reduces errors. Gathering data that varies and then trying to draw conclusions is extreme if you ask me. - - - Updated - - - This would be awesome! Sounds like exactly what we need to collect accurate data.
I like this idea. I'm already considering putting and old computer in the garage to act as a conduit for archiving dashcam footage, and have a server in the house pull through it. Making that server grab vehicle data over REST is on the list as well.
As part of my work at Plug In America, I've been planning to do a Model S battery study, like I've done for the LEAF and Roadster. I didn't think it was important to start this soon, but I could accelerate my time table if there's enough owner interest. I expect to release results from the Roadster study in the next month or two.
One word of caution: battery measurement appears to be changing still in some of the later software releases. My rated miles will jump up by 3-4 miles 10 seconds after plugging in, and when charging completes, it will drop by 3-4 about 15 seconds later. I've reported it, but it may throw off your degradation.
I've got no problem with the script approach and will chip in on the manual labor of dealing with the data as well.
In order to do this, we would have to be able to pull logs from the model S, like the roadster. Going by the rated or ideal miles after a charge is pretty useless at this point in time, because there are just too many variables that the firmware takes into account in order to arrive at those displayed numbers. The Rest API is a little better, but provides very little information compared to the log dump on a roadster.