Big data supporting sustainable mobility in smart cities

Transportation is responsible for 26% of Green House Gas (GHG) emissions at European level, ranging from 12% to 63% depending on the region. The contribution of the CO 2 to the GHG emissions is between 9 and 26%, being the second contributor in importance and the one most related to the transportation sector. In detail, road transport is credited with 72% of the overall sectorial emissions, presenting again a large variety of results in the different regions, from 10.8% to 95%. This variety of contributions depends on the sustainability of the mobility solutions provided in each region, which are significantly supported by Information and Communication Technologies applied to the transportation sector (named Intelligent Transportation Systems) as well as by Big and Open Data within the Smart Cities framework. Technological advances have been lately attributed with an increase both in the quality and quantity of mobility-related data, as by changing the way people perceive their everyday behavior, being in a constant state of sharing , they have turned citizens into active players of the data collection process. Luring them with the provision of services at real-time (in a variety of fields – from vehicle routing suggestions to traffic conditions updates), they have managed to engage people in a constant exchange of information, regarding both the individual and the environment. Smart phones and other devices offering “internet on the go” have succeeded in overcoming the naturally-set temporal and geographical limitations, making it possible for users to be connected at any place and time, thus rendering them to online information transmitters by sending and receiving valuable information to/from the content and services providers. Yet, the challenge of producing the best possible end-products out of these big datasets is twofold; on the one hand there is a need for developing algorithms able to filter, validate and process big amounts of data (almost) at real time, while on the other, there is a constant need for developing new applications and services for providing innovative and advanced traveler information services and traffic management schemes based on these data and processing capabilities. Within this paper, a framework for data collection, filtering and fusion is presented together with a set of operational tools in an effort to validate, analyze and highlight the added-value of big data. The framework is applied in the urban network of Thessaloniki, Greece, where two types of data sets are considered: probe data and traditional traffic data. The probe data set comprises of individual objects’ pulses (smart devices, navigators, etc.) tracked throughout the network at constant and pre-defined locations (“stationary” probe data collection) or during the whole trip of an “object” that continuously generates pulses (“dynamic” probe data collection). The traditional traffic data set is composed of data derived from inductive loops, cameras and radars. Both “networks” are implemented and operating in the city of Thessaloniki since 2013. Additional applications of the data sets are include mobility patterns identification, origin-destination matrices generation and validation, traffic flow estimation, route choice models development, macroscopic traffic and microscopic simulation models calibration as well as road hazards detection. The paper concludes with operational results regarding the above processes, highlighting key parameters, advantages and limitations of the use of probe data in urban regions. Finally, a discussion on the value of data is provided in order to promote open data schemes, which will considerably benefit the data industry.