Unpublished Manuscript ________________________________________________________________________________ SPEEDING IN TUNNELS: A COMMENT ON MANSER AND HANCOCK (2007) Author Francesco Biondi PhD student at University of Padova francesco.biondi@ymail.com Riccardo Rossi Massimiliano Gastaldi Claudio Mulatti Victims of excessive speed are numerous on European roads: according to the European Union, every year more than 1 million car accidents occur in Europe, leading to 100 deaths a day (ERSO, 2008). As reported by Stephan and Zimmermann (2008), a reduction in speed of only 3 km/h would save about 5-6 thousand lives every year. In light of these observations, many researchers have attempted to devise techniques to induce drivers to reduce speed by manipulating visual/perceptual factors such as vertical and horizontal road signs (e.g., Daniels et al., 2010; Gates et al., 2008; Godley et al., 2004; Jamson et al., 2010; Vieira-Gomes & Cardoso, 2012). Since optic flow represents, among others, an important source of information for speed perception while driving (Kemeny & Panerai, 2003), researchers often attempt to affect drivers’ actual speed by manipulating road geometry. For instance, in the studies of Agent (1980), Denton (1980) and Godley et al. (2000), the authors induced significant reductions in speed by adding transversal pavement markings at an increasing spatial rate to the road surface that affected drivers’ perceived speed. A similar rationale has been also considered by Manser and Hancock (2007) in their experiment in which, by the use of a driving simulator, drivers drove through a tunnel of 480 meters in length with the task of maintaining a pre-determined speed (about 80 km/h). At the entrance of the tunnel, the speedometer ceased to operate so that participants could not rely on it to regulate their speed. The authors manipulated the perceptual stimulation (texture) within the tunnel by applying vertical stripes to the tunnel’s walls. The participants experienced three distinct visual patterns: vertical segments that decreased, increased, or remained of a constant width throughout the length of the tunnel. There also was a control condition where participants drove inside the tunnel with no visual pattern present. The results showed that the presence of a visual pattern applied to the tunnel walls successfully affected driving speed; in particular, with respect to the control condition, the vertical segments decreasing in width induced a reduction in speed of more than 2 km/h. This is an important finding - with clear implications for traffic engineers – which the authors accounted for in terms of what they had called ‘perceptual speed regulation’. At a constant speed, the vertical bars pass the driver at an increasingly higher rate, given that the bars decrease in width. This gradually increasing rate would be perceived by the driver as an increase in speed. Since the task is to maintain a pre- determined speed, a perceived increase in speed would result in an actual speed reduction.