Скачать с ютуб BMW M2 x Elasticsearch: G87 Drifting with Kibana (RAW FOOTAGE) в хорошем качестве

BMW M2 x Elasticsearch: G87 Drifting with Kibana (RAW FOOTAGE)

A while ago I posted a short where I set up a datalogger on the PTCAN bus. It averages a few hundred megabytes of data per day of driving. The first #drift event of the year was last Sunday. So I swung down to collect some data. All packets are logged using can-utils' tool "candump" in a systemd unit file. This is then synced to ZFS (a .path unit file as well which checks to make sure the Pi is connected to my home's SSID) and a Python script on TrueNAS does some regex and uploads the packets in raw binary format to save space. Based on the data I can identify in SavvyCAN, I set up Runtime Fields in #Elasticsearch which extract ranges of bits from the raw bytes of each packet and allow me to graph everything out. Beauty of this approach is that it's lossless. I don't have to reupload/reindex every time I find new fields. So far I have 31 different signals of data decoded out of 280 unique #canbus IDs. This is a compilation of a few laps where I had my Dji Mic 2 set up inside, in the engine bay, and by the exhaust. Filmed with an Insta360 Ace Pro (1). Car is a 2024 G87 #M2 with 2025 M4 CS Transmission and 2025 M2 DME firmware for more boost and faster shifts. Drift Analyzer is only used to tighten the rear diff up. You can see it likes to unlock from time to time regardless; but it works. I got 23 laps out of a set of 255 Kenda KR20 (200tw) in the rear using a garden sprayer with ice in it after each lap. A bit excessive vs. other cars but they don't weight 4,000lbs with a driver. Oh well, I get more laps than them lol. Fronts are 285 ExtremeContact DWS-06+ which started to chunk apart (as a STEER TIRE) after just one lap. Do not recommend - switching to PS4S asap. Cheers.