MOVESMART addresses the problem of providing time-dependent route planning profiles in large-scale urban traffic networks by integrating the use of public transport, supporting electro-mobility and car-sharing/-pooling, exploiting an extensive ad hoc traffic monitoring infrastructure and a traffic prediction mechanism for foreseen future incidents, based on a novel crowd sourcing platform.
This exploitation of fixed and ad hoc sources of real time traffic sensing information provides proper alerts for emergent (either reported, or predicted) incidents to the involved end users and appropriate contingency plans for the predicted/reported disruptions.
To overcome the existing weaknesses for storing and maintaining traffic information, as well as in providing efficiently time dependent route plans, MOVESMART will employ a novel approach, based on the creation and maintenance of time dependent pre-processed urban traffic information that will allow the fast response to route-planning profile queries. Due to the quite demanding computational requirements for the creation and maintenance of this pre-processed information, our choice is to create an UTKB that resides in a cloud of urban traffic servers, and all the pre-processed data updates will be handled centrally by this UTKB.
The token is then passed to the portable navigation devices (either in-car devices, or smartphone applications), which submit route-planning profile queries to be handled consequently locally according to the end user’s particular preferences and the available options he or she is given. In order to deal with disconnection issues of the portable navigation devices MOVESMART will also provide a synchronisation procedure that will periodically feed the local devices with updates of time-dependent snapshots provided by the UTKB.
It also maintains the appropriate traffic-information data structures kept in the Urban Traffic Knowledge Base.
MOVESMART’s goal is to be able to respond in real time mobility – on-demand queries for efficient multi-modal route planning, both time-dependent, sensitive to aperiodic incidents (such as car accidents, sudden weather conditions, etc.) and forecasts provided by a Traffic Prediction Module, so that robustness and / or recoverability of the proposed route plans against reported but unforeseen disruptions is guaranteed.
To avoid the aforementioned weaknesses on traffic monitoring and digestion, MOVESMART will employ a novel approach that exploits the human-computer interaction via a crowd sourcing scheme mainly collecting traffic information from ad hoc traffic sensing entities. These ad hoc traffic sensing entities may be either in-car or in-hand portable devices (PNDs or smartphone applications) that will allow either periodic, or instantaneous on-route traffic reports, on behalf of the end users. The goal is to cover the entire metropolitan network, or at least the mostly used part of it, in a way that is complimentary to that of the fixed traffic sensing devices.
The ad hoc traffic sensing devices may report either periodic reports (e.g., average-speed reports) on particular road segments, or some alerts for emergent incidents such as a car accident that temporarily blockages of particular road segments, or a sudden weather storm that causes an unforeseen congestion. Coping with these unforeseen incidents, necessitates the consideration of dynamic networks, in which particular elements may suddenly become temporally unavailable, for a certain period of time. Finally each end user will also be able to submit post-route satisfaction reports with his or her own feedback on the quality of the route plans suggested by the system, in a like/dislike fashion.