GIS for Smart Parking
MetadataShow full item record
This thesis is about the phenomenon of parking in the city. Cities and especially bigger cities in The Netherlands, like Amsterdam and Rotterdam, have a relatively high number of visitors coming with their car every day. Because of scarcity of space in such areas this can lead to a parking issue such as not be able to park your cat at a desired location. Driving around looking for an alternative spot has negative environmental issues because of emmissions. The role of Geographical Information Systems (GIS) in this thesis, is that GIS are used to acquire parking data and to provide analytical- and scripting tools. GIS have been applied in this research field before, however, the emerging availability of open and realtime parking data gives new opportunities in analyzing parking dynamics. The aim of the thesis is getting insights in the dynamics of local parking systems. Which trends can there be identified? These trends are then put into models, which can be used in a system that predicts parking occupancy at a certain moment in time. The benefit of such system is that it triggers actions to be taken to prevent cars for driving to a full car park. The degree of emmissions are limited if cars are directed efficiently to available parking spots. Four different locations within the city center of Rotterdam are compared for exploring the dynamics in parking. Multiple models are developed depending on different variables: land-use and accessibility to public transport. Also the variables day, time and rainfall were tested on significance. With the exception of the latter these variables were significant and incorporated in the research. When comparing the models it turned out that the investigated land-uses showed different curves with peaks at a different moment during the day and having peaks at different days. Comparing these models to comparable areas in Amsterdam, resulted in the conclusion that each location has its own characteristics where different mechanisms are determing parking dynamics. Even if the land-uses are comparable, an area in general does not consist completely of just one land-use. Every person has its own motivations for parking their car, that does not mean someone is there for shopping or working but he or she can also park there because they are visiting a friend or park their car for continuing the trip by public transport to a different destination. The thesis also provides a system that can use prediction models to estimate when a parking area gets full in the near future. With this information, drivers can be directed to a different parking area where likely there are still available parking spots left. This limits emmissions that are produced by cars because it can prevent people ‘cruising’ around looking for a place to park. Parking data that are freely available are key for the developments of such prediction systems. Also technological developments for the collection of outdoor parking are important because outdoor parking data combined with indoor parking data gives a more complete view of parking dynamics in an area.