Multi-criteria analysis for evaluating the impacts of intelligent speed adaptation

Journal of Advanced Transportation, Winter, 2009 by D.B. Agusdinata, J.W.G.M. van der Pas, W.E. Walker, V.A.W.J. Marchau

Road safety is a policy priority due to the high casualties and costs associated with road accidents. Since speed is a major cause of road accidents, in-vehicle speed limiters or Intelligent Speed Adaptation (ISA), seems a promising solution. ISA implementation, however, is hindered by large uncertainties, for example about the impacts of ISA, the way users might respond to ISA, and the relationship between speed and accidents. Traditional Multi-Criteria Analysis (MCA) has limitations in handling these uncertainties. We present an MCA approach based on exploratory modeling, which uses computational experiments to explore the multiple outcomes of ISA policies (safety, emissions, throughput, and cost) across a range of future demand scenarios, functional relationships for performance criteria, and user responses to ISA. As an illustration, by testing the impacts of different ISA penetration levels on two driver groups, we show that when compliance with ISA is expected to be low, a policy aimed only at novice drivers outperforms other ISA policies on safety improvement.

1. Introduction

Transport policymakers around the world are faced with road safety problems and are attempting to improve road safety. Each year an estimated 1.2 million people are killed in road traffic incidents around the world (World Health Organization, 2004). In 2004 there were 1.3 million reported car accidents in the EU, resulting in 1.8 million casualties, of which more than 43,000 people died (Bialas-Motyl, 2007). Improvement of road traffic safety has become one of the major objectives for transport policies. Although various measures are taken, road deaths in the European Union in 2010 are estimated to be still some 32,500, so the EU objective set in 2001 of 25,000 road fatalities in 2010 is unlikely be met (European Commission, 2006).

1.1 Intelligent Speed Adaptation (ISA) as a potential solution

Many studies have established that speed is one of the major causes of road accidents (Hjalmdahl, 2004). Policymakers have been trying to reduce speeding using various measures, including driver-education campaigns, changes in infrastructure (e.g. placing roundabouts), and stricter enforcement of speed limits (e.g. using speed cameras). Although these measures have contributed to improving traffic safety, they seem increasingly unable to fully meet traffic safety objectives against limited resources. Therefore, additional measures are being considered that directly intervene in vehicle-driving tasks. In particular, systems that support the driver in not exceeding the legal speed limit, such as Intelligent Speed Adaptation (ISA), are of interest. These systems take into account local speed limits and warn the driver in case of speeding or may even automatically adjust the maximum driving speed to the posted maximum speed.

Since speeding is the major cause of traffic accident, the potential contribution of ISA to traffic safety is high. For instance, fully-automatic speed control devices are estimated to produce up to a 40% reduction in injury accidents (Varhelyi et al., 2001) and up to a 59% reduction in fatal accidents (Carsten et al., 2000). However, these estimates are likely to be high, since they are based on the assumptions that all vehicles are equipped with a perfectly functioning system, and that all drivers use the system in the way it was designed for.

The first ISA applications have recently entered the market. Speed limit information is being added to digital maps, so drivers can be warned about speeding through their navigation device. Future ISA can, in addition, intervene with the driving task. Systems using a haptic throttle have been tested in several field trials (e.g. in Lund (Biding et al., 2002), Ghent (Vlassenroot et al., 2007) and Australia (Regan et al., 2006)). When the driver exceeds the prevailing speed limit the throttle pushes back, providing an overridable resistance. Other trials have used a system that doesn't allow the driver to exceed the speed limit (e.g., in the UK (Carsten, Fowkes et al., 2000) and in Finland (Paatalo et al., 2002)).

So, the ISA technology is available and there is experience with using the technology. However, although expectations concerning the positive impacts of ISA are high, there still is a considerable gap between what is technologically possible and what has so far been implemented in practice (van der Heijden et al., 2002). The implementation of ISA is hindered by large uncertainties about the impacts of ISA, due to limited knowledge about such things as future transport demand, the way users might respond to ISA, and the relationship between speed and accidents (ADVISORS, 2002). For instance, the estimated contribution of ISA implementation to general transport safety policy goals often assumes a perfectly functioning system, that many vehicles are equipped with ISA, and that all drivers use the system in the way it was designed to be used. However, in practice, ISA systems have occasionally malfunctioned within tests ((Ministerie van Verkeer en Waterstaat et al., 2001; Regan, Triggs et al., 2006)), large scale implementation of ISA will take time, and some drivers have occasionally ignored the instructions when using ISA systems within these studies. As such, the real-world safety impacts of ISA implementation are uncertain.