Please add Fiat 500e (and a general EV question)

[carrrl]

My device arrives soon, but I noticed 500e is not listed in the “Car explorer” - can it be added as an option?

I see EVs require additional setup for power cycle messaging. Besides that I imagine I can configure each bus, add all the PIDs I’ve decoded, configure triggers and loggers, then it works like any vehicle?

Thanks!

[Peter]

Hi @carrrl,

Welcome to the community.

Please email support@autopi.io, they can assist you in creating the vehicle in the list.

best
Peter

[carrrl]

After speaking with support I wanted to document my experience with this vehicle.

• Fiat 500e is not OBD-II compliant (EVs aren’t required to be in USA) but AutoPi really wants it to be
• CAN bus detection fails with Internal RS232 transmit buffer is full due to high traffic (2,000+ msg/s) on CAN-C, but can be manually detected and added via CAN Analyzer
• Bus detection is not always “silent” and can cause error frames on CAN-C which then throws all kinds of cluster errors. CAN-B cannot be manually or automatically determined because it’s constantly messing with CAN-C - Why can I not just assign specific pins to busses/protocols?
• Power cycle cannot be configured as suggested for EVs. PID queries require extended headers, but UI is locked to the typical 3 characters. CAN filter creation is disabled (seemingly because bus cannot be verified/auto detected?)
• Since these two core functions are not available to me, the entire value proposition is lost, all other features hinge on that core functionality.

What are my options? I am basically left with an old RPi running a web terminal that wants to go to sleep every 5 minutes.

Is there hardware documentation so I can at least roll my own stack top to bottom? Raspbian, SocketCAN, ACC power would work just fine, surely that can be done with this device? I hoped AutoPi would save me time, but it seems not

[carrrl]

After further digging I cannot recommend this device to 500e users and issue a warning to everyone else. There are just too many blockers and red flags at this point. I’ve made more progress in a week with python and $70 in parts than nearly a month looking into AutoPi.

Issues:

• PID and CAN only work with 11-bit IDs. Full stop, this product is useless if you want to use extended 29-bit IDs.
• Pins 1/9 connect to our CAN-B, but AutoPi can’t handle that out of the box. At first glance we could potentially wire up an adapter cable to remap 1/9 to 3/11. “Vendor” pins should be an option possibly by use of internal jumpers.
• Documentation is light and what does exist is often outdated. Old terms and UI don’t always translate to the current product.
• This product is mostly aspirational - look at what you could build - but the software/hardware/support makes that incredibly difficult.
• The more I search, the more I find no answers or suggestions you contact the company to buy support and feature development.

Worst of all, this is such a fantastic opportunity for the company yet they seem to be absent. The foundation of this product is good. The hardware is easy, the software for all it’s faults could be great. I guess I expected more for 2-3 years in the market.

I could honestly see a market build around this platform. Aftermarket car stuff is typically overpriced junk that doesn’t work together. This device could have been a platform to unify it all for people who want to add “modern” features to otherwise basic vehicles. That would sell more AutoPi by actually delivering on the aspirations. People want a weekend project, not a full time job reimplenting stuff you sold us.