Evaluation Telraam2.0 and counting loops

Intent

The Mobility Company of Ghent undertook a pilot project to map traffic intensities with a relatively new method, namely Telraam. A first trial already ran from the end of 2019 to 2020. The conclusions of this trial can be found in an earlier evaluation report . In the meantime, a second trial started at the beginning of 2023 (with a new version Telramen2.0), within the Compair project. In Ghent, 15 Telrams are being deployed and tested in the Dampoort district within this project. The report below explains how a Telraam sensor works, how the pilot project was set up and how the new Telraam sensor was evaluated. The evaluation focuses on three important aspects, namely data availability, data quality and the data collection process.

Conclusions

It seems that in several streets Telraam gives a reliable indication of the total number of vehicles passing in the period that there is sufficient visibility. The vehicle classification does not correspond with the classification by counting hoses, which makes it difficult to determine whether and how much 'misclassification' occurs. If there is a deviation between the two measurement results, this can usually be explained by external factors.

• As already suggested, it is important to use the same measurement method for pre- and post-measurements. The Telraam data does seem to generate data in a consistent manner, which makes the data source seem suitable for trend analyses and pre-post analyses.
• Counting frame data seems less suitable for use as a short-term measuring instrument (e.g. one hour of data).
• The vehicle category that is most correctly estimated (by both counting tubes and counting frames) appears to be that of passenger cars or the number of motorized vehicles.
• It would be a (major) added value if the counting frame data could be filtered, so that unreliable measurement moments (e.g. due to sun impact or because a direction of travel is not counted) can be filtered out by means of, for example, control counts.
• It remains important to approach the data critically and to use control counts to determine whether the data from a specific device can be used for further analyses.

Based on the previous analysis, the counting frames seem to us to be particularly suitable for long-term analyses: is the total number of vehicles increasing/decreasing, how is the number of passenger cars and/or the total number of motor vehicles evolving... It certainly also seems an interesting avenue to investigate whether the outdoor variant can be used as a fixed counting point to monitor motorised traffic in and around Ghent.

Finally, a nice addition is the online platform that was made available within the Compair pilot. It makes it easier for administrators to get an overview of which counting frames are active within a network and what the status of the devices is.

For more information, we refer you to the report below.