I mean, you're attempting to reverse-engineer a complex electrical system. If you have confirmed that no simpler methods of activating the controller don't work, it is actually on on the larger battery and waiting to receive the CAN payload to turn on, then I think this is the most straightforward and safe option to follow.
You will need some sort of programmable microcontroller ideally with onboard CAN and a bus transceiver (ideally on a breakout board). I am personally using STM F103 (blue pill board) and TJA1050. It is possible to use a microcontroller without CAN, but then you will need a dual controller-xceiver adapter board that will talk to it via e.g. SPI.
The setup is pretty straightforward. Hook up bus to the adapter, hook up adapter to microcontroller. Software can be written in your embedded framework of choice - I recommend Embassy but Arduino will work fine too. You can look at my Votol repository for example communication with the controller and steal that code if you want.
All of that is relatively straightforward and safe as long as you do preliminary reading and research.
I would make a new thread specifically for your setup if you decide to try this, as it will likely generate a lot of replies during support, which would spam this thread unnecessarily.
