Counting lit streets in the winter/dark

I see in the overview that Telraam v2 can’t count in the dark. That makes sense. In the middle of winter, we have darkness from 16 to 09, so this would drastically reduce the effectiveness.

But what if the street is lit?

I have also noticed that my S2 stops counting when you can easily see cars and bikes. I think that the sensitivity of the sensor is not totally optimal, which surprises me a little compared to what you now find on cheap phones.

This is likely to bias the measurements as total cars are not normalised to the effective counting time.

Interesting.

At what time/darkness level does that occur?

For reference, a useful day/night visualization: Day and Night World Map

We never have “night” at this time of year, only nautical twilight, and it seems to be able to keep counting through that in a sense. I can’t vouch for the accuracy, though.

1 Like

I was wondering if anyone has figured out an easy way to “normalise” data due to the changing length of daylight during the year?

Our S2 shows a steady decline in traffic, which seems to correspond to shorter days, but it would be great to be able to easily adjust for the number of hours of daylight.

1 Like

To address your question @balchen , here in Copenhagen we have a Telraam that hangs in a lamp store that counts to late in the night (sometimes until 1 at night) which is kinda a cool coincidence. It is a small shopping street with lanterns on the other side as well. We did not select for this but is kinda cool. The accuracy also has been fine.

You need a lot of light for this and the device also shuts down after midnight, because even in summer it becomes difficult for most places outside of the Arctic Circle.

@dave.anderson.ie If you pull the data with the Telraam API t you could integrate this data into your BI tool and normalize it by dividing traffic counts by daylight hours. This should help you account for variations due to changing day lengths. I am not a datawizard, but with some of our new AI friends, pivot tables and MS Powerbi it is very doable.

1 Like