Case studies

Use big data to detect parking behaviour

The Kortrijk case focuses on integrated parking management. In Kortrijk, more than 1000 on-street parking spots are equipped with a dedicated parking sensor, which registers the status of the parking places. Besides the 1000 on-street parking sensors also open and live data is available of all the public parking garages and terrains.

In the city centre, it is only allowed to park your car for a half-hour on one of the 1000 so-called ‘shop & go’ parking places. Parking sensors monitor the actual status, and the parking attendants can monitor real-time where cars are in overstay. The same sensors can be used to monitor the behaviour in the other parking zones (with a maximum stay of 2 hours) in the city. By using a statistically relevant sample, the current parking policy in the city centre can be scrutinized in more detail.  

The Kortrijk case will provide insight into how advanced sensors can be used for policymaking and policy measuring. By combining the relation between different sensors, AI can be used to predict the most efficiënt parking policy to support a livable city and to support the local economy.

Challenges encountered:

The PoliVisu Kortrijk case leans on years of experience with (automated) parking management of the Kortrijk City parking company Parko.

The first challenge we will encounter is to get a better insight into the parking behaviour in the 2 hours zone. It’s important to check the number of cars in overstay (parking more than 2 hours) and the unpaid parking places.

A second challenge is to get an idea of the parking pressure and the impact on overstay.

The third challenge is to define and test a prediction model for parking pressure based on historical data.

Because the Kortrijk case is brand new, it isn’t possible at the moment to explain how we did overcome the challenges. A major challenge is setting up the measurement equipment in the 2 hours zone and finding a way to predict the effect of changes in the parking policy. 

Stakeholders:

  • Parko (Kortrijk city owned parking company)
  • City of Kortrijk
  • Informatie Vlaanderen

Actions steps:

Defining clear research questions (policy preparation) and defining of a first scientific approach to have a qualitative sensor test sample.

Lessons learned:

At this stage, it is still early to identify lessons learnt. The existing day-to-day live data parking occupancy map gives already an insight into the parking pressure during the week and year, in zones with many overstay offences. This information was already used to study parking pressure and to elaborate where parking zones are needed and what parking regime make sense to be applied.

Kortrijk has a policy of parking data ownership and is one of the few Flemish cities who own the data of all the public parking in the city. The parking data ownership is part of the exploitation contracts with the private parking companies running those public parkings.

Outcome impact:

Kortrijk is at least in Belgium, the city with the most comprehensive parking data. By using parking sensors also on-street parking can be well monitored. Kortrijk has today good insights because of their ownership on parking data that is in many cities not available because of a lack of knowledge about the exact number of parking places, the occupancy of these places and a lack of data of public parking. 

The impact can be further improved by extending the current dashboard containing historical and live data with a simulation engine (what-if analysis) and trend analysis (predictive) engine.

Links:

 

In the future, Kortrijk wants to obtain the following datasets:

  • Data from the on street 2 hour parking zone sample sensors (when these sensors are available);
  • ANPR data to measure the in- and outgoing traffic.

 


Related articles

About The Author

Leave a Review.

You must be logged in to post a comment.