I've been thinking a lot lately about solar power, and specifically about the supposed need for expensive and heavy steel structures to hold panels in place, at least in some applications. And I've also been thinking about electric vehicles, and about how electric engines are much more efficient than internal combustion engines, even before considering whether the source of energy is renewable.

And the idea occurred to me: why not just put solar paneling on the roof of the vehicle itself?

Now, obviously a couple of standard size panels aren't going to provide the peak power demand of a car engine. But the EV does have a battery. The numbers I found are that typical panels can generate on the order of 200 W/m^2 in full sun [1], while typical EVs use about 200 Wh/km [2]. Thus, given that there are a few m^2 of usable space on the roof, several hours of exposure would cover many kilometres of travel. You could drive to work, leave the car in an open-air parking lot, and perhaps capture enough energy to cover the drive home. In general it would at least cover part of the electricity demand, adding up to considerable savings over the long run.

I'm imagining a system of small (say 10cm to a side) tiles on a lightweight rack, with a simple feedback control system able to tilt the tiles to face the Sun over the course of a day. The whole thing would be integrated into the roof of the vehicle, with a glass or polycarbonate cover, such that there's no need to worry about snow, wind etc. on the panels themselves. It does of course imply forgoing a sunroof.

The numbers should actually work out better than the average case, I'm thinking, since the panels could directly generate the car's native DC voltage — skipping any inverter and avoiding any significant transmission losses. The panels just need to be able to charge the battery that's already there (out of necessity).

In the past I've thought about putting roofs over parking lots with panels on top. Using the car roofs instead would distribute the costs and savings to individuals and avoid expensive infrastructure (that might also cause political or regulatory issues).

Are there major problems with the idea that I'm not thinking of? Is anyone seriously working on something like this?

[1]: e.g. https://www.energydawnice.com/solar-panel-output-per-square-meter/ ; this reads to me like SEO-optimized slop, but so does everything else I can find, and it all seems to roughly agree anyway. Although from what I know about insolation and the reported efficiency of modern panels, I imagine it should be possible to do a fair bit better than that.

[2]: https://ev-database.org/cheatsheet/energy-consumption-electric-car