Anyone else using a Raspberry Pi to log Victron data locally without VRM?

I've been running a Victron MPPT 100/30 and a SmartShunt on my narrowboat for about eight months now, and whilst the VRM portal is handy enough, I really don't want to rely on a mobile data connection for logging — especially when I'm moored up in the middle of nowhere with no signal. I've got a Raspberry Pi 4 sitting spare and I'm convinced there must be a decent way to pull data locally over USB or VE.Direct and store it myself.

I've had a look at the official Victron Venus OS large image for the Pi, and that seems like the obvious route — basically turns the Pi into a Cerbo GX on the cheap. I've got it running on a test card and it seems stable, but I'm not 100% sure whether I can have it log to a local InfluxDB/Grafana stack and still push to VRM when I do have signal, or whether it's one or the other. Has anyone actually got that dual setup working reliably?

The other option I keep reading about is bypassing Venus OS entirely and just reading the VE.Direct serial data raw with a Python script, then feeding it into Grafana myself. Feels like more work but also more control. I'm happy enough with Python but I've never done anything with InfluxDB before — is the learning curve brutal or is it fairly straightforward once you've got the basics?

CoastalBoater | 847 posts | ⚓ Pembrokeshire

@JulieEvans great project! I've been running almost exactly this setup on my bilge keel for two years now. One thing worth mentioning that I haven't seen flagged much — if you're on a narrowboat with limited 12V stability, power the Pi from a proper DC-DC converter rather than a cheap USB adaptor. I lost about three weeks of logged data from SD card corruption before I sorted that out.

I use Venus OS Large on a Pi 4, connects to both devices over VE.Direct, and I feed the data into InfluxDB with Grafana dashboards. Completely offline, looks brilliant on an old tablet mounted in the cabin.

The SignalK integration is also worth a look if you want to tie in other boat instruments down the line. Happy to share my config files if useful.

PanelSteve | 1,203 posts | 🚢 Somewhere between a lock and a hard place

@JulieEvans been doing exactly this for two years on my boat. The short version: venus-data-logger on a Pi Zero 2W, pulling from the MPPT and SmartShunt via

Jonno25 | 312 posts | 🔋 Yorkshire Dales

@JulieEvans yes, been doing this for about a year on my off-grid cabin setup with a MultiPlus and a couple of MPPTs. The key bit people often miss is setting the correct product_id in the dbus-serialbattery config if you're pulling data from a SmartShunt specifically - caught me out for weeks.

I'd also strongly recommend InfluxDB + Grafana over anything else for local visualisation. Slightly more setup overhead but the dashboards are genuinely brilliant once you've got historical data building up. Run the whole lot on a Pi 4 with a small SSD rather than SD card - saves you the headache of card corruption down the line, which on a boat with vibration is very much a when not if situation.

Happy to share my docker-compose file if useful?

Jonno13 | 578 posts | ⚡ Narrowboat & static van hybrid

@JulieEvans brilliant setup to go for! One thing worth adding that I haven't seen mentioned yet — if you're using a Venus OS device or a Pi running Venus OS Large, you can enable the built-in Node-RED instance which makes building custom dashboards dead straightforward without needing much coding knowledge. Pair that with InfluxDB and Grafana running locally on the same Pi and you've got beautiful historical graphs completely offline.

Also worth checking out the dbus-mqtt or dbus-serial-battery paths depending on your exact hardware config. The Victron community forum has some cracking guides for the serial connection side of things.

What are you planning to use for the actual display — dedicated screen on the Pi, or just accessing it over your boat's local WiFi from a tablet?

Simon1988 | 847 posts | ☀️ Off-grid since before it was trendy

@JulieEvans great timing on this thread! I'd add that if you're running a Pi for local logging, it's worth looking at Venus OS Large on a Pi rather than just pulling data via serial/USB. Turns your Pi into a proper GX device - you get the full Victron ecosystem locally without needing VRM at all. Pair it with InfluxDB and Grafana and your dashboards are genuinely better than VRM anyway, plus everything stays on your boat's local network.

One narrowboat-specific tip: power the Pi from a decent DC-DC converter rather than a cheapo USB charger. I lost weeks of data when mine browned out at a dodgy mooring with low voltage. A Pololu or similar rated down to 7V input sorted it completely.

AndyRobinson | 1,204 posts | 🔧 Garden office + motorhome

@JulieEvans the piece nobody mentions until it bites them — serial vs VE.Direct vs Bluetooth matters enormously here. Your MPPT 100/30 and SmartShunt both have VE.Direct ports, so grab two VE.Direct-to-USB cables (genuine Victron ones, the clones can drop packets) and run venus-data-logger or full VenusOS Large if you want Grafana dashboards baked in.

I run a Pi 4 in my garden office pulling from a Multiplus and two MPPTs simultaneously — stores locally to InfluxDB, visualises in Grafana, zero VRM dependency whatsoever. On the narrowboat I'd specifically look at signalk integration too since you've presumably got other marine instruments worth tying together.

One practical note: power the Pi from a regulated 5V USB rather than direct 12V — the voltage sag when your starter motor kicks can reset it mid

SunnyViking | 312 posts | ⚡ Garden office & narrowboat

@JulieEvans been running exactly this on my narrowboat for about a year — Pi Zero 2W connected via USB to a SmartShunt, using venus-data.tgz scripts to pull data into InfluxDB and Grafana. Completely offline, no VRM dependency whatsoever.

One thing I'd add that nobody's flagged yet: power draw matters on a boat. The Zero 2W idles around 0.8W versus a full Pi 4 at 3-4W. Over a winter that difference adds up when your panels are barely breaking a sweat. Worth factoring into your load calculations before you commit to hardware.