Urban Driving

Urban traffic presents a far more heterogeneous spectrum of scenarios than freeway traffic, particularly regarding road users, infrastructure, buildings near the roadway, and dynamical timescales. Well-designed experiments involving assistance systems require comparable scenarios and user characteristics in order to perform proper data analysis and draw reliable conclusions concerning needs. The sub-project “Urban Driving” will establish a unified platform including internal data standards with defined driver assistance scenarios for all the UR:BAN projects. There are at least three important sets of variables determining a “driver assistance scenario”: first, the characteristics of the assistance of information system in question, such as reliability, intervention characteristics, and system limits; secondly, the driving task level (navigation, maneuvering, stabilization) addressed by the assistance or information functionality; finally time requirements and operating load on the driver (permanent, intermittent, mode-specific). These three dimensions of variability will require close coordination of the partners, right from the beginning of the project.

In UR:BAN, considerable effort will be invested to obtain valuable data regarding critical or hazardous traffic situations. Thus, it is important to coordinate these activities so that multiple projects can all utilize these data resources; measurement and evaluation standards also need to be established to provide optimal quality in interpretation and assessment of results within all projects. To this end, previous work within the EU projects FESTA and Euro-FOT as well as from SimTD will provide valuable input: these projects have developed data standards that can be implemented by different partners with reasonable technical effort. By adaptation EURO FOT specifications to the needs of UR:BAN rather than starting from scratch, coordination of standards can be accomplished quickly. These standards include sampling rates, video formats, synchronization parameters, and situation coding. Standardization is essential and saves valuable experimental resources, for example, when it is necessary to collect additional data to extend or validate a small experimental sample.  Coordinated scenarios are also required in the sub-project “Human-Machine Interaction for Urban Environments” for the planned fusion of different functions into an integral use concept. Integration of different systems is often hindered simply because function developers use triggering conditions with indeterminate timing, so that it becomes impossible to pre-process data and coordinate information systematically for the user. A statement such as “The information should be presented during approach to the intersection” could cover a time interval up to one minute, depending on scenario.

By the end of UR:BAN, systematic specifications will be revised in light of the functional developments in other UR:BAN projects, and quantitative characteristics defining acceptance, usability, controllability, and traffic impact will be available for assessment of the relevant assistance scenarios.