Active Development

Navigation your
EV actually deserves.

evOS is a vehicle intelligence and navigation platform for electric cars that ship with navigation too dumb to respect their own battery. Real EV routing, live vehicle telemetry, and offline maps β€” running on hardware you control.

EV-native
Routing engine
Real-time
Vehicle telemetry
Offline
Map data
Open hardware
No cloud dependency
How it works

Three layers, one system

evOS runs entirely in the car. A small embedded module reads vehicle data directly, a compute unit handles routing and intelligence, and a touchscreen puts it all in front of the driver.

πŸ”Œ
Vehicle Interface
Reads live data directly from the car β€”
battery state, speed, temperatures,
charging status, and more.
β†’
πŸ–₯️
Intelligence Core
Runs on a Raspberry Pi in the car.
EV routing, navigation logic,
telemetry logging, map serving.
β†’
πŸ“±
Driver Interface
Touchscreen HMI β€” navigation,
battery overview, media controls.
Also available on Android.
πŸ—ΊοΈ
Offline Maps
Served locally, no internet needed
πŸ”
Address Search
Fast local geocoding
🎡
Media
Spotify, no phone required
πŸ”§
Remote Admin
Manage the system without touching the car
Capabilities

Built for EV reality

Generic navigation doesn't understand your battery. evOS does.

⚑

Topology & speed-aware consumption forecasting

Energy use is forecast per road segment based on elevation change and expected speed β€” not just distance. Uphill, downhill, motorway, urban: each costs what it actually costs.

πŸ”‹

Automatic charger selection

evOS finds and inserts the right charging stops along your route automatically β€” factoring in your current SOC, arrival margin, plug availability, and charge speed. No manual planning needed.

πŸ“‘

Live vehicle telemetry

Reads directly from the car's own data bus. Pack voltage, current, state of charge, temperatures, and vehicle state β€” streamed in real time to the interface.

πŸ—ΊοΈ

Offline-first vector maps

Map data served locally β€” no internet dependency in the car. Full country coverage fits on a small storage card. Works in tunnels, remote areas, abroad.

πŸ“Š

Drive & charge session recording

Every trip auto-recorded: GPS track, SOC curve, power, speed, altitude, road type. Review sessions with summary stats or stream the raw data for analysis.

πŸ”¬

Deep battery insight

Goes well beyond the dashboard SOC number. Individual cell monitoring, pack health tracking, real-time power flow, and charge state estimates the OEM never surfaces to the driver.

πŸ”’

10 independent trip counters

Track distance, energy, and consumption across 10 separate counters β€” reset them independently. One for the current charge cycle, one for the trip, the rest for whatever you want to measure.

🏎️

Fast long-distance routing

Route calculation that stays fast on long trips while remaining accurate for local roads. Energy cost is modelled per road segment β€” not just distance.

🎡

Spotify β€” no phone required

Full Spotify playback running directly on the in-car computer. Album art, track controls, and auth β€” all inside the HMI.

πŸ”§

Over-the-air updates

Push firmware updates to the vehicle interface module directly from the HMI. No laptop, no cable, no workshop visit.

Foundation

Built to last

evOS is production software, not a prototype. The architecture is designed to adapt to other EV models with minimal integration work.

βš™οΈ

Systems-grade core

The intelligence core and vehicle interface are written in Rust β€” chosen for reliability, performance, and the ability to run safely on embedded hardware without compromising the compute layer.

πŸ–₯️

Native, cross-platform UI

The driver interface runs natively on Linux (Raspberry Pi touchscreen) and Android. One codebase, two targets β€” so the same HMI works on a windshield tablet or a dedicated display.

πŸ”—

Clean internal interfaces

Every component communicates through well-defined contracts. Swapping in a different vehicle's data source, a different map provider, or a different display is a targeted integration β€” not a rewrite.

Development Platform

Born on the smart 451 ED

Every feature in evOS was built and proven on a real vehicle β€” one that needed it.

πŸš—

The smart fortwo 451 Electric Drive is a capable little EV with one serious problem: its OEM navigation and vehicle software are essentially non-existent. No range-aware routing, no battery insight beyond a basic charge bar, no trip data worth keeping.

evOS was built to fill that gap β€” and in doing so, became a platform that can do the same for any EV whose manufacturer decided software wasn't worth the investment. If your car has a data bus and a driver who deserves better, evOS can work for you.

Field Testing

Trip reports

The best spec is a real trip. Here's where evOS gets tested in the wild.

First view of the Adriatic coast near Poreč
πŸ‡¨πŸ‡­ β†’ πŸ‡­πŸ‡·

Maiden voyage β€” Croatia

Spring 2026 Β· Switzerland β†’ Croatia

The first long-distance run with evOS fully operational. Alpine passes, motorway corridors across Austria, and the Adriatic coast β€” with charge stops planned and logged in real time.

822 km
Distance
9
Charge stops
0%
Lowest SoC
Read report β†’
The smart ED charging near Imst on the way home, alpine peak behind
πŸ‡­πŸ‡· β†’ πŸ‡¨πŸ‡­

The way back β€” over the Arlberg

2–3 May 2026 Β· Croatia β†’ Switzerland

The return leg β€” home a different way, past Salzburg and through Germany to dodge the Brenner. So uneventful we stopped logging the numbers and started feeling smug. With one charge stop to go, that turned out to be a mistake.

882 km
Distance
119
Wh / km
1,791 km
Round trip
Read report β†’
For Manufacturers

The problem isn't unique to one car.

Across the EV industry β€” passenger cars, vans, trucks β€” drivers are working around software that wasn't built with electric mobility in mind. evOS exists because one driver got tired of waiting for someone else to fix it.

I've been driving electric for over a decade β€” more than 400,000 km across Tesla, Mercedes, and smart. I know what good EV software feels like, I know what bad EV software costs you on a long trip, and I know the difference between a feature that looks good in a brochure and one that matters when you're at 12% in a valley with the next charger 40 km away.

Building evOS required going further β€” deep integration with the vehicle's own battery management system, decoding proprietary data, understanding how the car actually behaves rather than how the spec sheet says it does. That kind of work isn't something you learn from documentation alone. You learn it by doing it, on a real vehicle, on real roads.

If you're an OEM β€” passenger EV, electric van, electric truck β€” and you want navigation and vehicle intelligence that is genuinely integrated rather than bolted on, I'm interested in that conversation. Not an aftermarket workaround. Something that starts from the vehicle's own data, speaks the driver's language, and actually helps them get where they're going with the energy they have.

The commercial EV space is solving this in the open right now β€” drivers and independent developers filling gaps that manufacturers haven't prioritised. I'd rather work with an OEM to do it properly.

400,000+ km
driven electric
10+ years
as an EV driver
πŸš— Passenger EVs
🚐 Electric vans
🚚 Electric trucks
🚌 Fleet & transit
Let's talk
Transparency

Built with AI assistance β€” openly

AI-assisted development is the new normal. This project uses it and doesn't hide it.

evOS was built by one developer across embedded hardware, a systems-level compute core, and a native UI β€” a stack that would have taken years solo. A significant share of implementation was done pair-programming with AI tools, which handled boilerplate, scaffolding, and debugging while I drove architecture, integration, and product decisions.

The result is a more capable system, shipped faster, with more breadth than would otherwise be realistic for a side project. What I bring: the system design, the hardware integration, the product judgment, and the stubborn willingness to drive a small electric car across the Alps.

πŸ€– Claude Code β€” agentic coding sessions
⚑ Cursor β€” inline AI editing
Get Involved

Interested?

Whether you're an EV enthusiast, a potential acquirer, or just want to talk shop β€” reach out.

🀝 Acquire the project

evOS is an active side project with no paying customers β€” yet. If you see potential in an EV-native navigation platform with a working hardware integration and a real vehicle deployment, let's talk.

Send an email

πŸ’¬ Talk shop

Interested in the architecture, the EV integration approach, or just want to swap notes on squeezing range out of a small electric car β€” happy to chat.

founder@interware.ch