Sten Ruppe

Risk minimal routes for emergency cars

The computation of an optimal route for given start and destination in a static transportation network is used in many applications of private route planning. In this work we focus on route planning for emergency cars, such as for example police, fire brigade and ambulance. In case of private route planning typical quantities to be minimized are travel time or route length. However, the idea of this paper is to minimize the risk of a travel time exceeding a certain limit. This is inspired by the fact that the emergency cars have to reach the destination within a legal time. We consider mainly two approaches. The first approach takes into account relevant information to determine the weight, i.e. the desirability of certain edges of a graph during the minimization procedure. One possible risk factor to be aware of would be a suddenly jammed single-lane road on which the emergency car has no chance to make use of the benefits of the siren for instance. The same holds for full-closure situations and railroad crossings. We present a catalogue of risk factors along with an appropriate algorithm for practical route planning in emergency situations. The second one takes into account a weekly updated set of probe-vehicle data for each minute of the week along with data of current travel times. Comparing those travel-time data allows calculation of the associated risk for traveling certain edges of a route in a road network. We expect our algorithm to be a major advancement especially for destinations that lie outside the typical region travelled weekdays. In this case the automatic route planning naturally goes along with an additional gain of time.

Ranking Of Alternatives For Emergency Routing On Urban Road Networks

Routing on urban road networks for emergency cars is an application of Dijkstra’s algorithm with relevance in everyday-life. Since distances in urban transport are rather short it is computationally possible to calculate many paths and compare them afterwards. This paper uses Dijkstra’s k-shortest path algorithm in order to calculate shortest and fastest paths and finally finding an ordering of alternatives for multi-criteria routing. The solutions are displayed in criterion space and the Pareto front is identified. Routes are ranked according to the normalized weighted-sum method. Obviously, the more alternatives there are the more possibilities for the emergency car to circumscribe traffic jams. Therefore ‘close alternative routes’ are taken into accounts that share a certain fraction of nodes with one Pareto optimal route. To those bundles of routes a ranking is assigned that may serve as recommended action for the driver.

Analysing Velocity Data For Emergency Cars At Urban Crossroads

Concerning disaster management with respect to urban transport, one vital and lifesaving part of this sector is the driving of emergency cars. In urban German areas, firemen have to reach their destination within a legal time of eight minutes within 90% of the cases of emergency. In order to achieve that, they are equipped with special right-of-way when activating their flashing blue light. They can drive with increased speed, they have priority, they can use emergency lanes formed by other cars and they can pass over red traffic lights. However, although they are equipped with these special rights, they have to be careful that no accidents happen and that nobody is injured. Consequently, at each intersec-tion they typically have to reduce their speed. Therefore, it is difficult to predict for a given fire station the actual region that can be reached within the legal time. Obviously, crossings form a major bottleneck along their route. Therefore, the present paper considers real data for emergency cars at crossroads. The velocity data of emergency cars passing the crossroad are compared in case of driving (1) with blue-flashing light and siren, (2) only blue-light driving, and (3) without any signal. The results are a first step towards a realistic prediction of the trajec-tories of emergency cars in case of emergency and help to identify bottlenecks that should be avoided if possible. The results can be used for special routing for emergency cars and are of importance for typical cases of emergency as well as large-scale applications of disaster management.