All Categories Telraam v1 How to validate your own Telraam data (V1)

How to validate your own Telraam data (V1)

If you want to validate the data your Telraam data, this article will guide you through the steps to perform a manual validation of the accuracy of the counted number of road users. In order to perform a validation of the speed, you would need specialist equipment like a (laser) speed gun, so we will ignore this option in this article.

Step 1 - Preparation

1️⃣    Before starting the data validation process, it is important to understand several definitions and concepts, such as 'precision' and 'accuracy' of a Telraam, how the validation is done correctly, what can be causes of inaccuracies, ...

 

👀 Therefore, first read the following article,  👀 
"
PRECISION, ACCURACY, AND VALIDATIONS OF THE ORIGINAL TELRAAM SENSOR"

 

2️⃣    Print the validation manual count sheet that you can use to conduct manual counts (click here to download the pdf)

3️⃣    Choose the day and time to perform the manual count. Note that this should not be during twilight hours and try to avoid very bright times of day with strong shadows.

Instead of watching the road for the requisite time (15 minutes for S2 or 1 hour for V1) you could choose to record a video of the road segment, and do this at your leisure later.

Step 2 (V1) - Perform the manual count

4️⃣    Perform the manual count for 1 hour starting and ending as close to the hour mark as possible.

Step 3 - Confirm that the data quality of your Telraam was good

5️⃣    Go to your dashboard

6️⃣    Go to 'Overview in detail'

7️⃣    Look for the day and time for which you performed the manual count

8️⃣    Confirm that the data of this time interval was 'good quality' or 'poor quality'

 👍🏼    If 'good quality' --> that means your Telraam had a good uptime (v1) or good visibility (S2). You can move on to step 4.

 👎🏼    If 'poor' quality' --> that means your Telraam did not have enough uptime or light conditions were poor. In this case the data from this interval may not be of high enough quality to make a comparison with your manual count. Unfortunately you will have to start over.CleanShot-2022-07-28-at-16.13.37@2x.png

The timestamp label in this overview is the initial hour. So 16:00 is the period from 16:00 to 17:00.

Note: for S2 users without access to a Pro Dashboard, please contact support@telraam.net to get the readings for the specific 15 minute period (please give us the segment ID and the exact timestamp required)

 

Step 4 - Make the comparison

9️⃣    The first figure you need to compare is "total road users"

Inaccuracies (when the manual count data does not match the Telraam data) can be caused by:

➙   The Telraam sensor cannot see the whole road surface

➙   Obstructed or merged objects

➙   Traffic jams with stopped traffic for long periods

➙   The Telraam sensor is in an overly sun-lit window

 🔟    Secondly, you can compare the numbers by modality.

Inaccuracies (when the manual count data does not match the Telraam data) can be caused by:

➙   Very low or high camera view angle.

➙   Long, strong shadows

➙   Obstruction, or tight grouping of road users

➙   Atypical vehicle composition

Read more in the previously mentioned article about these causes: "PRECISION, ACCURACY, AND VALIDATIONS OF THE ORIGINAL TELRAAM SENSOR"

 

Was this article helpful?

Thanks for your feedback!