Open kezarjg opened 2 months ago
We actually have a bulk
endpoint that's available for exactly this sort of thing. I'll update the API docs to include it, but this is definitely possible.
https://documenter.getpostman.com/view/7396339/SWTK5a8w#4f77e0b8-aa60-4a6e-83df-3c3f7fe19aec
The format is fairly simple, and the innermost "tlm_list" elements are identical to the tlm field of the current format.
OK. I'll take a look. Do the entries need to be sorted by time? I've been getting some odd jumps in my trip data (like when I might have connectivity problems)
Assuming the timestamp (UTC seconds) on the entry itself is correct, it should be added to the database and displayed properly. We do discard out of order data for calibration though, as we need to go sequentially.
It is definitely better to send the points in sorted by time though, oldest to newest.
https://documenter.getpostman.com/view/7396339/SWTK5a8w#4f77e0b8-aa60-4a6e-83df-3c3f7fe19aec
The format is fairly simple, and the innermost "tlm_list" elements are identical to the tlm field of the current format.
Does the API key need to be authorized for bulk uploads? I'm receiving the following error; 401 Unauthorized Key
@jason-abrp My current thinking is to send bulk telemetry every 60 seconds. Will that cause any issues with ABRP?
That should be fine, but maybe a little choppy on the displayed SoC in-app while driving since it could then lag behind by up to a minute.
OK. I'll do some testing with every 10 seconds.
@jason-abrp - Any suggestion what to do with data when no connection is available. Is there value to ABRP to send the data 'later'? Does it help with average consumption or charging data for vehicles?
Sending the data later is good for filling in missing segments in your drive history, but it's not really used by the calibration process. Up to you if you want to include that.
I would like to propose an enhancement to support collecting data at a high frequency and sending it in batches. Specifically, I am interested in understanding if there is an API call to support:
This would be particularly useful in scenarios where: