Smart parking in IoT scenarios based on data integration with MobiVoc

published on by Annette Weilandt

In the EU research project bIoTope, we are reviewing and extending MobiVoc for use cases concerning smart routing and parking for vehicles. Our goal is to provide an integrated dataset containing parking data of several cities that can be used by consumers in IoT scenarios.

In order to achieve this goal we map and enrich multiple open datasets that provide parking information from the cities of Brussels, Helsinki and Lyon by using the tool eccenca Corporate Memory. The resulting dataset is published for other consumers within the IoT environment of bIoTope. The challenge we have to overcome when integrating the data sources is that the datasets provided by the three cities contain different information and use different models to describe the data. This means that although all datasets offer parking information, this information is expressed differently in detail. For example, whereas one city only indicates the total number of available and occupied parking spaces, another city breaks down this information into parking spaces for certain vehicle types (car, mortorbike, etc.) and indicates the number of vacant spaces available. Another problem is that the data itself as well as the data descriptions are given in different languages (French, Finnish, Dutch) and are provided in different file formats.

But despite the usage of different models the datasets describe in principle the same contents i.e. facts that are relevant for finding a parking space. It is like saying the same thing with different words. So, MobiVoc wants to provide a standardized model for this domain. The individual elements of the datasets are mapped to the MobiVoc vocabulary to obtain a harmonized dataset that includes the data of the various parking data sources. The real datasets of the cities are very useful to further extend the vocabulary and to adapt it to the requirements of the domain. With such a basis, MobiVoc is made ready for an implementation in real use cases. As a second step, the harmonized data is enriched by other RDF datasets like dbpedia. The new integrated and enriched dataset provides more benefits for consumers since it contains more information than the single data sources and follows a uniform semantic data model.

Within the bIoTope project this dataset will be offered as a data service in an O-MI-Node of the IoT environment. Other consumers like connected cars can use this data source to improve navigation and routing in vehicles.

Integration and publication of parking data in bIoTope

Internet of Things, MobiVoc, Mobility, Open Mobility Vocabulary, Parking, bIoTope Project