We build our own vehicle camera system: cameras on the car, a computer in the car, detection running live at 15 frames per second while you drive. One pass produces a pole inventory, the assets on each pole, and the vegetation along the corridor — as data, not as hours of video nobody will watch.
The driver's screen from a real drive: poles drop off the map as the cameras capture them, the table fills up with what was detected on each pole, and the same street is rebuilt in 3D.
▶ Open the live demo
Starts only when you ask: it uses the graphics card.
Illustration of one vehicle pass: corridor scan, front and rear camera, detections as they come in.
The cameras run at 15 fps and the detector runs on board, so what leaves the vehicle is the finding, not the footage. A day's survey is megabytes, not petabytes.
A single forward camera means driving every street twice. Two cameras cover both sides in one pass and roughly halve the driving time.
At normal traffic speed, with the route planned so no street is driven twice for nothing.
We analyse poles, so we charge for poles: a yearly fee per pole, not per square kilometre, not per image, not per hour of processing. Your cost is known before the first drive and does not move.
Here is our proposal. Look at the city drives on this site first — and notice what they were recorded with: an ordinary action camera on a windscreen, the kind anyone can buy for a few hundred dollars. Nothing else. Send us a few streets of your own and we will run the same analysis on them, free of charge. It is a test, not a survey: it shows you what comes out, on your own network, before you decide anything.
First run free, one area per organisation, up to 30 minutes of footage. Get in touch and we will send the upload details.
Be clear about which one you are looking at, because it decides what to expect. One is a test you can run yourself this week; the other is the survey you buy.
Most of the cost of a camera survey is not the camera. It is the video you keep and the cloud you rent to look at it afterwards. Our system removes both.
We analyse poles, so we charge for poles. Not per square kilometre, not per image, not per hour of processing. The number is known before the first drive.
Once the camera is on your vehicle we do not count passes. After a storm, twice a season, every month — the annual fee per pole does not move.
Every analysed pole lands on the server. Both sides see the same number of poles covered, so there is nothing to argue about at invoicing time.
| Europe | North America | |
|---|---|---|
| Per pole, full survey, per year | €1 | $1.1 |
| Average span assumed | 40 m | 150 ft (≈ 46 m) |
| Poles per unit of line | 25 per km | 35.2 per mile |
| Per line length, per year | €25 / km | $38.7 / mile |
We bill per pole. Line-length prices are averages at the spans shown; denser lines cost more per km, sparser lines less. Invoicing in EUR or USD. Unlimited passes included.
See it working before you ask for a price.▶ Open the live demoIn our experience urban areas often have no consistent pole records at all — and where records exist, they are not much help in the field. In unfamiliar streets, in traffic, with a van parked in front of the pole, no crew can keep track of which pole was actually captured. So the screen does it.
Open the driver's screen and drive it yourself →
No stock footage and no mock-ups: every number on the city pages comes from a camera that actually drove that street, and every claim that is only an estimate is marked as one. Where we have not measured something — a vegetation height without a 3D reconstruction, for example — the page says so instead of showing a number. The defect list we detect today is the one you can see on those pages; we extend it to whatever list your own inspections use.