Evvu | Electric Vehicle adoption decision‑support App

I started from human bias. My preliminary research (interviews and surveys) and the academic literature on EV adoption pointed the same way: users tend to overestimate their actual needs. That set the design problem. I had to measure a driver's real mobility and hand it back in a form that was hard to argue with.

The core method was a data‑driven digital twin. The two options people already had, self‑reporting and dealer consultancy, both inherit the user's own bias: you cannot answer a perception problem with a perceived answer. So I modelled an EV's behaviour against the driver's real trips instead. The app reads everyday GPS journeys, then replays each one through a virtual EV the user picks, showing live whether that trip would have been covered. Compatibility becomes something you watch happen on your own routes.

From concept to product, I worked in Lean Startup cycles, sharpened during the I3P incubation: build the smallest testable version, put it in front of real drivers, read what their data said back. I stayed close to the build rather than throwing specs over the wall: understanding what the GPS data could and could not support, what was cheap to test and what was expensive to get wrong, let us argue the right trade‑off on the same evidence.

Business design ran in parallel, because a tool that helps people but cannot sustain itself never moves a market. I framed Evvu as Software as a Service: free for drivers, while energy providers, car dealers and public administration licence the engine as an API and receive aggregated mobility data in return.

The strategic logic was to move the decision from persuasion to proof. Evvu does not try to talk anyone into going electric, it lets the car prove itself against the driver's own life. When the choice rests on personal evidence instead of a salesperson's promise, going electric stops being a leap of faith.

Using Evvu is simple. The user downloads the app, grants the location permissions, and picks the electric model they are curious about. Then the user drives as they always do, in the car they already own. In the background, Evvu builds a digital twin of that EV and runs it against the user's real mobility data, trip by trip, in live simulation. Next time the user opens the app, they see their own driving needs next to how the simulated electric car performed, and whether it would have kept up. The simulation also allows the user to compare the two vehicles in terms of emissions and fuel consumption.

The B2B layer exposes the same engine as an API for energy companies, dealers, manufacturers, and public administration, who embed it and get aggregated mobility data back. That aggregated data was the most strategic asset of all, especially for energy companies. Knowing the real routes that EV‑ready drivers already travel tells a utility exactly where new charging stations would pay off, on the corridors that matter rather than on a guess. Place the chargers there and you answer the last doubt a borderline driver still holds, where will I charge, before they even ask it. The same behavioural data that nudges one person toward an EV is what builds the infrastructure to convert them for good, which closes the loop back to the environmental goal that started the whole thing.