Accueil  >  GERI  >  ANIMATIC 

Spatio-temporal Data Collection and Analysis for Supporting Driving Behavior and Mobility

Scientific Coordination: Corinne Brusque (DS), Nour-Eddin El Faouzi (COSYS/Licit), Latifa Oukhellou (COSYS/Grettia)

Program and presentations 

Scientific aims

  • to create a scientific forum for discussion and exchange aiming at better understanding mobility and driving behaviors stemming from the collection and analysis of spatio-temporal data from the observation of movement;
  • to promote methodological Advances with inter-research units collaboration on topics related to mobility and driving behaviors. This multidisciplinary  purpose can be served by co-supervision of internship and/or master thesis;
  • to facilitate setting up consortia and common proposals within national and european calls and tenders;
  • to value these exchanges by editing collective volumes summarizing the knowledge capitalized by this IREG's participants during the completed seminars/workshops.

    ANIM@TIC IREG (Ifsttar Research and Exchange Group) aims first to bring together researchers from several research units (GRETTIA, LICIT, LESCOT, LIVIC, LEPSIS, LVMT, DEST, IM/EASE, UMRESTTE, MA, MACS, ...) working on spatio-temporal data related to driving behavior and mobility. These data is collected either from sensing devices or through observational mechanisms, e.g. surveys, origin-destination (OD) tables.
    Note that mobility patterns and driving behaviours are traditionally analysed through human and social sciences framework. The emergence of Information and Communication  Technologies (ICT) as well as the advent of new observations and tracking capabilities have boosted of sizable spatio-temporal data. The amount of collected data contribute to leverage the development of novel multi-disciplinary approaches linking human and social sciences with engineering and computer science.

    With respect to mobility, and more specifically intelligent mobility and traffic engineering, the progress of tracking devices and communication (cell phones, satellite localization, mobile internet ...) makes it possible to develop a fairly new and innovative approaches for traffic modelling. Indeed, a large amount of high resolution mobility data becomes available, such as vehicle kinematics, travel time and/or vehicle trajectories. In addition, data related to  other forms of mobility such as bicycle sharing systems or pedestrians are being collected thanks to GNSS tacking systems (e.g. GPS). Tools for processing spatio-temporal data are needed in order to better understand mobility patterns of travellers and goods as well as the use and performance of transportation systems based on a set of performance indicators.

    Regarding driving behavior, observation and analysis of driving behavior in different scenarios (e.g. normal vs accident situations) are handled from different angles with various approaches. Within this topic, various transport modes are considered (motorized modes such as automobile and motorcycles, or non-motorized ones).
    Data collection are usually based on dedicated experiments  with equipped vehicles continously monitoring and recording driving data in actual driving situations, including normal situations as well as critical ones such as near-crash and crash situations. Thus, the recorded data provide a flow of information on the vehicle dynamics, on the driver's actions on the vehicle, and also on the driver's behaviour and the driving environment.  These data, classically recorded for limited trajectories, are now observed and recorded unobtrusively in a natural setting of driving over a long period of time. Among the naturalistic observation approaches, one can mention Naturalistic Driving studies (NDS) or Field Operational Tests (FOT).  These data offer interesting possibilities for driving behavior investigation in terms of knowledge, accident risk, eco-driving, or design of driver assistance. Nonetheless, developing innovative methods of observation, data collection and processing is still an open question which deserve much attention in the future.

    Ifsttar is involved in these issues through several projects, among which European FP7 projects such as: FESTA, DACOTA (for the methodological aspects of the observations in natural situations of driving), INTERACTION, EuroFOT (including observations on instrumented vehicles), 2besafe (including observations on instrumented two-wheelers), OPTIMUM, ECOSTAND and NEARCTIS network of Excellence. Researchers from various disciplines  are facing scientific and methodological questions that deserve to be debated.  The ANIM@TIC IREG is sought to be the right plate-form for such debate and interdisciplinary open forum.
    Indicative list of topics (not limitative) that could be addressed by ANIM@TIC

      • Observation protocol and experiments design
      • Spatio-temporal data processing and analysis approaches
      • Generalisation methodologies on how  results obtained on a particular sample can be extended over the whole population
      • Using the near-crash or crash to predict the relative risk
      • Legal issues related to the tracking the people and mobiles
      • Sharing ressources and Database access.

      There will thus be both academic presentations and discussions with practioners aimed at fostering interactions in this important field which lies at the boundaries of civil and social engineering, computer science and ICT technologies.

      International Workshop December 5, 2012 - Marne-la-Vallée (France)
      Spatio-temporal data mining for a better understanding of people mobility. The Bicycle Sharing System (BSS) case study


      Expected results in 2012

      • Kick-off workshop and a set up a steering Committee
      • Competencies and skills inventory with Ifsttar and the scientific and technical Nerwork of minitry of Transport
      • Identification of topics that should be considered within ANIM@TIC
      • Organization of two quarterly workshops.