The parking spot search problem (PSSP) is a problem millions of drivers face each day. It poses an economic cost as well as an environmental burden. Smart cities oer new ways of assisting drivers to reduce the time spent looking for a free parking spot and the distance between the nal parking spot and the actual destination. We compare three fundamentally dierent approaches to the PSSP in extensive simulations on a grid network with randomly generated routes and varying vehicle densities: First, and as a reference, a nave random approach that does not use any smart hardware. Second, a global approach where communication of atomic parking spot availability data ows through a central server that can be reached by all vehicles in the network. Third, a completely distributed approach where vehicles gather such information themselves and only share it with their geographical neighbors. Our results show that such smart approaches do reduce search times and remaining distances signicantly. The centralized approach performs best in all scenarios. However, it poses the strongest assumptions, from a theoretical perspective as well as on actual infrastructure, and its deployment would be much more expensive than a decentralized solution. Such a distributed approach achieves nearly as good results with a local memory that only stores information about 5 parking spots. It is therefore an important option to consider when trying to improve driver satisfaction in urban areas.
Masterprüfung mit Defensio, Reinhardt Moritz Bastian
21.11.2018 11:00 - 12:30
Organiser:
SPL 5
Location:
Besprechungsraum 4.34
Währinger Straße 29
1090 Wien
1090 Wien