AI CAN Bus Sniffer Data Pack
Join our newsletter to get your 'AI CAN sniffer data pack'!
This pack lets you reproduce the showcases from our AI CAN bus reverse engineering article using the CANsub, python-can and a Claude Code skill. It contains the following:
- Raw CAN data from various cars with matching OBD2 reference signals (speed, RPM)
- Our Opel Astra raw CAN data + dashboard video recording to test the 'vision' workflow
- Ready to use with the open source can-bus-reverse-engineering-skills Claude Code skill
The pack lets you replicate the exact showcases from our AI CAN bus reverse engineering article. With the included data and the open source Claude Code skill, you can:
- Reverse engineer proprietary speed/RPM signals using the OBD2 reference data (the 'log file' workflow)
- Test the 'vision' workflow on the Opel Astra data, where the dashboard video acts as the reference signal
- Compare your results against the decoded outputs shown in the article
Install the open source can-bus-reverse-engineering-skills Claude Code skill (see the README), then point Claude at the relevant CSV log file and reference data from the pack. The skill handles the rest, as walked through step by step in our AI CAN sniffer article.
For the vision workflow, you provide both the raw Opel Astra CAN trace and the matching dashboard video so the skill can extract a reference signal via Optical Character Recognition (OCR).
The showcases are based on our CANsub.2 CAN bus interface and python-can. To capture your own car data, you also need an OBD2-DB9 adapter (and optionally our contactless adapter if your OBD2 connector exposes no proprietary CAN).