This project focused on the development of a knowledge infrastructure, computational models, and user applications that allowed access to real-time information about the state of transportation-related resources as well as predictions regarding their future state. A pilot service that exemplified the usage potential of available data will be provided to citizens for making public transportation more efficient and pleasant to use and to policy-makers as a decision-support tool.
The project was organized around the following specific objectives:
- Acquiring and parsing data that described the state of transportation-related resources (e.g. bus and train locations, cell phone traces, road sensors, GPS tracking, weather, emergency events, and census data);
- Developing a data fusion engine that combined the data to extract from them additional information including predictions;
- Developing a model-based data fusion engine (with simulation capabilities) that modelled a particular transportation network along with the behaviour of travellers within it;
- Developing a web-based service that enables different applications to access the data collected;
- Developing an interactive multi-modal, multi-criteria route-planning application that would exemplify the capabilities of the data fusion engine and provide support to the travelling decisions of the public;
- Developing a geographical modelling and visualization module of city dynamics for use by planners and service providers.
CityMotion was the first Portuguese project that combined a mixture of data from several (often competing) sources. For the first time in Portugal, the project was advancing “reality mining” from heterogeneous (and differently owned) data sources.
CityMotion has spawned a second phase proposal for a large national project in the mobility area (TICE.mobilidade, QREN funding program) involving a large number of private companies and the universities of Coimbra, Porto and Minho.
Prototype of the Portugal Brisa A5 motorway online laboratory is being developed.
Initial testing of various online-calibration algorithms within DynaMIT were carried out and preliminary results on the quality of traffic estimation and prediction, as well as computational performance were obtained and analysed. Furthermore, a parallel calibration strategy, which leads to dramatic reductions in algorithm running time, was designed and implemented.
- Innovative visualizations of mobility patterns for Lisbon;
- Development of density matrixes, O-D matrixes and clustering results involving different mobility modalities. Study of relations between mobility and the urban space;
- Adaptation of DynaMIT to Brisa’s A5 motorway, to develop a “short-term” traffic prediction tool.