EUROPE - BRAZIL COLLABORATION OF BIG DATA SCIENTIFIC RESEARCH THROUGH CLOUD-CENTRIC APPLICATIONS

EUROPE - BRAZIL COLLABORATION OF BIG DATA SCIENTIFIC RESEARCH THROUGH CLOUD-CENTRIC APPLICATIONS

Press [ esc ] or close+

Connected Societies

EUBra-BIGSEA: driving innovation in Massively Connected Societies with high impacts on business and society through common priorities for Europe and Brazil


Anyone living, working in or visiting a large city knows something about how hard it can be to get around, whether by car or public transport. EUBra-BIGSEA is working on two related applications that help tackle congestion problems and improve urban transportation planning.

BIGSEA not only investigates speed, vehicle flux, traffic disruptions, main origin-destination routes for cities based on the day, time and area covered, but also covers the bigger human side, such as feelings and topics emerging from historical or recent data that is associated with specific places, stress caused by traffic, landmarks, the presence of green areas, weather conditions and their effects on people moving about a city.

 

End Users

Citizens and urban planners will benefit the most from the dual implementation of the application for massively connected societies from two different perspectives.

Mobile and web interfaces will enable citizens to query route options available ahead of a planned journey. This helps improve travel time, including reducing stress levels, while making the journey as pleasant and interesting as possible, depending on the day, hour and location. Route recommendations will draw on long-term historical data to identify trends, and short-term data focusing on recent and relevant events. State-of-the-art and novel predictive models from data science also play a role in making the recommendations.

Urban planners benefit from an informative view of mobility in the city, with data on state of play, trends, and impact of events. The application will use data mining descriptive models and enable urban planners to carry out interactive investigations, also benefitting from notifications and alerts. These innovations will help urban planners both operate and plan transportation systems more effectively.

 

Tackling multi-faceted challenges with BIGSEA innovations

EUBra-BIGSEA addresses many different challenges that range from privacy, quality, ethical, political, and social issues to technological and engineering complexities.

From a computational viewpoint, implementing the use case from two different perspectives requires non-trivial resources and services from a cloud platform. On top of this, large volumes of data will be continuously acquired, processed and stored. But the data also needs to be pre-processed in different ways to extract value for higher level functionalities and mining. Models are also required for descriptions and predictions on the transportation system that meet the right requirements. This is compute-intensive and time-consuming. What’s more, understanding and predicting the multifaceted attributes for a route is challenging, as we need to consider events like accidents and football matches that are external to the computing system.

This is where the innovations from BIGSEA come into play. BIGSEA is bringing together navigation algorithms already used in transportation systems with novel data mining models that understand and predict origin-destination characteristics and visualization, with a focus on the need for better city transportation systems.

BIGSEA is leveraging big data and four data sources to create data mining models that tackle the challenging, multi-faceted attributes of routes through a city:

  • Detailed stationary georeferenced data about the city is integrated with other data sources.
  • Three big data sources cover the dynamic aspects of the system: GPS location of all buses and user cards in the city, data output of fine-grained models of weather, historical and real-time data posted on social media about the city and its locations.
  • Integration and continuous processing of this diverse and large volumes of data helps detect patterns, trends and outliers (observations at abnormal distances) in the behaviour of transportation systems.