Time: Monday 2nd November 2009, 13:00
Place: Aalborg University, Selma Lagerløfs vej 300, room 02.12
In partial fulfillment of the terms for obtaining the Ph.D.-degree, Dalia Tiešyte will give a lecture on the following subject:
“GPS Data Management with Applications in Collective Transport”
A number of applications in areas such as logistics, postal services, cargo delivery, and collective transport involve the management of fleets of vehicles that are expected to travel along known routes according to either fixed or flexible schedules. With the spread of centralized, real-time position tracking of vehicles, transport-related systems are now able to provide up-to-date journey-related information to a variety of users, as well as store the position-related data for further analysis.
Due to road construction, accidents, and other unanticipated conditions, the vehicles’ travel times deviate from the expected schedules. At the same time, there is a need for the infrastructure surrounding the vehicles to continually know the actual status of the vehicles. For example, anticipated arrival times of buses may have to be displayed at bus stops. It is a fundamental challenge to maintain this type of knowledge with minimal cost, and to provide the real-time information to the interested parties, such as managers of the systems, and the clients or passengers.
This thesis addresses the problems related to the development of a real-time vehicle location management system, with the focus on vehicles traveling on pre-defined routes. This involves real-time vehicle location tracking using wireless communication, management of on-line and off-line trajectories, and prediction of the future status of the vehicles when their movements are restricted to given routes and when they follow schedules with best effort. The thesis proposes novel tracking, trajectory data recovery, vehicle trajectory similarity search and indexing, trajectory data analysis with a focus on predictability, and generalized, adaptive prediction methods.
Extensive empirical evaluation of the proposed, as well as existing, methods is performed using both generated and real data. The real data was collected from public buses in Denmark using the GPS system. The thesis includes extensive empirical analyses of this bus trajectory data and evaluations of various prediction algorithms using this data. We show, that the prediction problem is highly data-dependent. Differently from the existing prediction methods, we aim to generalize the problem, and to allow adaptation to the data and context.
Members of the assessment committee are Associate Professor Vladimir I. Zadorozhny, University of Pittsburgh, USA, Associate Professor Yannis Theodoridis, University of Piraeus, Greece, and Associate Professor Simonas Šaltenis (Chairman). Professor Christian S. Jensen is Dalia Tiešyte’s advisor.
All interested parties are welcome. After the defense the department will be hosting a small reception in cluster 3