The strip of ground between two railway rails has never had a use.
Ballast, drainage, the space a maintenance crew walks along, and nothing else.
On a branch line in western Switzerland it is now a power station, laid flat, level with the sleeper tops.
A train passes over it every hour or so.
Wheels turn a few inches above the glass.
The panels have not moved and have not needed replacing.
Why anyone would put a panel at the worst possible angle
Every installation guide says the same thing. Face the sun, angle toward it, and match your latitude roughly.
Lying dead flat throws that away. At this latitude a flat panel loses about a tenth of what a properly tilted one would collect over a year.
So the case for doing it has nothing to do with the physics of the panel and everything to do with the ground underneath.
That strip is already owned, already cleared, already fenced, already drained and already reached by a maintenance access route.
Nobody has to buy a field, argue with a planning authority or build a mounting structure on virgin land.
You give up ten percent of the yield and you skip the entire land problem, which is the constraint that actually holds solar back in a mountainous country.
What is actually lying between those rails
The panels sit in aluminum frames clipped to the sleepers, and every one of them lifts out.
That matters more than it sounds. Track has to be tamped, ground and inspected on a schedule, and anything permanently bonded to it would be an obstruction.
A purpose built train rolls the strip out and picks it back up, which is closer to laying carpet than to construction.
The glass is tougher than a rooftop module and carries an anti reflection filter, so a driver approaching at speed is not dazzled by the line ahead.
Brushes mounted on a passing train sweep the surface, which is how a system with no access road stays clean.
Sensors are built in, so the operator is reading the installation continuously rather than walking the track to check it.
What the first year actually returned
The site is a 328 foot stretch at Buttes in the canton of Neuchatel, on a line that runs a passenger service through a valley.
Forty eight panels, 18 kilowatts of capacity, energized on the twentieth of May in 2025.
By the anniversary more than 11,000 trains had passed over it.
The installation produced over 16,000 kilowatt hours in that year, which is the annual consumption of three or four Swiss households.
That figure is the point of the whole exercise, because it is what the design predicted. The strip hit its target rather than falling short of it.
The transport regulator set the trial at three years. One year in, the results are strong enough that a permanent installation is now considered likely, and the founder is pushing to shorten the timetable.
What 328 feet of track cannot tell you
One valley branch line is not a network.
The line is single track, lightly used by European standards, and runs through cool mountain air that keeps the panels near their efficient operating range. A busy main line at 125 miles an hour is a different vibration and a different thermal problem.
Nor does one year say anything about what a decade of frost, grinding dust and heavy tamping does to a frame.
Cost is the other silence. No audited figure for the installation has been published, and the economics of a system laid and lifted by a specialist train are not obvious.
The precedent everyone reaches for is discouraging. France paved a Normandy lane with panels and dug it up after it never made half of what was promised.
The difference is that a road surface takes the load directly and a rail strip never does, because the wheels ride on steel a few inches above it.
What happens if it does scale
Switzerland has about 3,300 miles of railway outside its tunnels.
Covered end to end that would theoretically produce around a billion kilowatt hours a year, enough for roughly 300,000 households and about two percent of national demand.
Two percent is a real number and a modest one.
The queue behind it is longer than the pilot. A cooperation agreement with the French operator was signed in February, an Italian pilot is in discussion, South Korea has approved a two year trial, and Dutch, Chinese, Indian, Singaporean and Indonesian firms have all made contact.
The next step technically is to stop sending the output to the local grid and feed it straight into the railway’s own traction supply, which is where a railway’s demand actually sits.
This began as a question about whether solar panels could survive under a train, and the original trial was framed exactly that way.
Eleven thousand trains later the survival question is closed. What is open is everything financial.
Read the whole thing?
Get the week's signal, not the noise
Our sharpest reporting on energy, climate and nature — free, once a week.
