redragonx / can2rnet Goto Github PK

Thanks for your great work! I am currently working on an invacare power wheelchair with Linx electronics. I want to receive data and control the wheelchair remotely. This is a work in the scope of my phd thesis, but it is also a personal project, because I have 2 sisters who use wheelchairs 🙂

So far I am simply using arduino mega, interfacing with the wheelchair bus, using the chip MCP2515 (Can High, Can Low to SPI interface). I'm already able to read some frames, but I still can't decrypt them totally. I would like to send some of these messages to the bus to decrypt which frame can wake up the wheelchair, or make some engine work, but the joystick module shows an error and the system turns off.

Now I was wondering, because I saw in a project that use can2rnet that joystick is induced to error and to be able to communicate with the power module in its place.... is this the only way? because I wanted my joystick to work normally as it controls the state of the batteries and charges for example.

Also, do you think wireshark could help me better decipher some of this information? I think I would receive the same data, but maybe with less losses.

Thank you so much!

What do you think guys? @ysshah I think we should.

I am thinking that we can make a modularized plugin base system that allows for adding input systems easily. Like the April Tag sensor/section of smart wheels. And then make it easy to add other subsystems.

Ideally, I want to implement Bluetooth Beacon tracking via phone or any other BT device. This would be easier to use around people and doesn't rely on camera tracking. The chair PI would have multiple BT adapters to track the signal and to figure out direction/distance. And with additional external sensors to detect the environment around the chair. This is what can2RNET was originally designed for anyway, so I would like to start on this soon. :D

We had to remove the "-overlay" from these lines in /boot/config.txt

dtoverlay=mcp2515-can0-overlay,oscillator=16000000,interrupt=25         
dtoverlay=spi-bcm2835-overlay

so it became

dtoverlay=mcp2515-can0,oscillator=16000000,interrupt=25         
dtoverlay=spi-bcm2835

for it to work. Maybe it was due to a Raspbian update?

To fix the wait_rnet_joystick_frame function in JoyLocal.py, we can use Python's select module as described here to set a timeout on the recvfrom method.