Why does my Telraam need time to calibrate?

Attention: Telraam needs time to calibrate. During the registration process, you will probably have noticed this disclaimer. But what does it mean? And why does Telraam need time to calibrate?

Why the need for calibration?

In practice, the disclaimer points to the fact that Telraam data is not reliable in the first hours and days after installation of the sensor. During the first few days of counting, your Telraam lacks sufficient data to establish a difference between cars and large vehicles (known to us geeks as the “cut-off”). This is why, in the first days, Telraam uses a general rule to establish the cut-off. Because this rule is general and not adapted to your specific Telraam, the resulting data is usually incorrect.

How long does Telraam take to calibrate?

Over time, as your Telraam keeps counting, data reliability will automatically improve. This will take around a week.

Please note that in winter, when days are shorter, Telraam collects less data and the calibration period takes longer: up to three weeks.

Cars and large vehicles

In some cases, even after a calibration period, Telraam does not correctly distinguish between cars and large vehicles. The number of cars can be underestimated and the number of large vehicles overestimated, or vice versa. The sum of both vehicle types, however, is usually correct.

At this moment in time, the car/large vehicle split point is set roughly around pickup trucks. It is, however, possible that large cars (SUVs, pickups,…) are counted as large vehicles and small, low-slung vans are counted as cars. Click here to learn more about how Telraam categorizes vehicles.

In summary, we ask you to be conscious of the fact that Telraam’s large vehicle count data does not have the same accuracy as the data on cars.

oke so much for the bad recognition of cars and large vehicles. a week then 3 days is not enough. But what about the excuberant number of bikers?

Hi Diana,
Do you have the feeling your Telraam counts too much bikes?
A vehicle is classified as a two-wheeler based on the average observed fullness (the red part on the picture) and the axisratio (the arrows on the picture):
CleanShot 2022-09-22 at 12.33.12@2x

An example of misclassficiation could be: a group of cyclists are being counted as a car, because they ride really close to each other and the algorithm mistakenly classifies them as a car because the average observed fullness and axisratio of the group of cyclists weren’t close to the values of an individual cyclist, and closer to the values of a car.
A pedestrian with a stroller could with the same logic mistakenly be classified as a cyclist.
A scooter will be counted as a two-wheeler as well. I hope this explains a bit why and when you can expect this kind of misclassifications?

Curious to hear if anyone else experiences the feeling that their Telraam counted too many two-wheelers? Or maybe to other way around?

I reset my telraam device prior to your reply, but after my comment. Nowadays all is well.

1 Like

More useful content here for @mounir.djama as well - hope this helps to clarify