1 Assessment of Speeding Profiles and Safety Margins from Tangent to Curve by means of Driving Simulation Eleonora Papadimitriou 1* , Stergios Mavromatis 2 , Dimosthenis Pavlou 3 , George Yannis 4 1 Department of Transportation Planning and Engineering, National Technical University of Athens, 5 Heroon Polytechniou str., GR-15773 Athens, nopapadi@central.ntua.gr 2 Technological Educational Institute of Athens, School of Civil Engineering and Surveying & Geoinformatics Engineering, 2 Agiou Spiridonos Str., GR-12210 Athens, Greece, stemavro@teiath.gr 3 Department of Transportation Planning and Engineering, National Technical University of Athens, 5 Heroon Polytechniou str., GR-15773 Athens, dpavlou@central.ntua.gr 4 Department of Transportation Planning and Engineering, National Technical University of Athens, 5 Heroon Polytechniou str., GR-15773 Athens, geyannis@central.ntua.gr Abstract This paper presents a novel definition of drivers’ safety margins reflected in speed profiles on a tangent to curved road design. These safety margins are based on a vehicle dynamics model, which is implemented to assess the speed variation at impending skid conditions from tangent to curve on the basis of several parameters. This model returns the theoretical speed-distance curve corresponding to the driver’s maximum safe speed and acceleration when utilizing the outmost of the available vehicle horse power. On the basis of actual vehicle speed profiles, the model also returns the respective curve for the actual speed-distance i.e. the utilized share of vehicle horse power, which reflects the driver’s safety margin. Data from a driving simulator experiment are used to test the proposed methodology, explore driver’s speed profiles and the parameters affecting drivers’ safety margins. The results suggest that drivers’ safety margins towards the examined curve are considerable, with the majority of the drivers using less than 55% of the available vehicle horse power. Drivers can be grouped into “aggressive”, “moderate” and “conservative” speeding behaviour, each group exhibiting distinct initial speed, “breakpoint” distance and acceleration / deceleration patterns. Higher initial speed is positively correlated with more aggressive driving i.e. lower safety margins. On the contrary, a higher safety margin was associated with earlier deceleration before the curve. The proposed approach yields a continuous and objective assessment of driver speeding behaviour from tangent to curve and the related safety margins, both for individual drivers, as well as for groups of drivers with similar speeding patterns. Keywords Speeding; Tangent-to-Curve; Safety Margin; Vehicle Dynamics model; Simulator experiment. 1. Background and objectives Design consistency is acknowledged to be a key element for road safety [1]. This is achieved by avoiding abrupt changes of critical alignment elements that may result in erratic driving maneuvers and eventually crashes, so that successive elements of the road act in a coordinated way to enhance safety [2-5]. A critical element is the approaching between tangents to horizontal curves [6 - 8]. Design consistency is typically assessed on the basis of the operational speed [6], and substantial differences between operational speeds or between design and operational speeds in successive design elements, indicate poor design consistency. In this context, the “safety margin” is often defined as the difference between the driver’s speed and the design speed, and is intended to express the degree of safe speeding behaviour at curves or from tangent-to-curve. However, only the examination of the operational speed variation between the curve and the preceding tangent seems inadequate. The reason is that most researchers use spot speed values along the approach from tangent to curve, and vehicles’ acceleration/deceleration is extracted assuming either a linear relationship between the measured spot speed data or a linear regression analysis based on the curvature [e.g. 7,10]. In general during vehicle motion on tangents, especially long ones, the drivers do not maintain a constant speed; they usually tend to accelerate their vehicles [6,8]. However, at some point before entering a curve, the drivers adjust (decrease) their speed accordingly.