Ambulance drivers are required to make decisions on which route to follow under time constraints because, in such cases every second is counted to save a life. Timmons, 2007 indicated that “Decision makers are susceptible to cognitive biases when operating under stress, i. e. , high workload, time pressure, and information ambiguity”, pp4. A cognitive bias is defined by Haselton et al. (2005) as the tendency to search for information and alternatives that prove their preconceptions and to discount information that disproves their preconceptions.
Kowalski-Trakofler et al. (2003) carried a study to discuss human judgement and decision making under stress. The authors selected recent literatures and carried out a field work to discuss the affect of stress on emergency responders. They examined coping with stress under time constrains on expert emergency teams. It was concluded from this research that “stress restricts cue sampling, decreases vigilance, reduces the capacity of working memory, causes premature closure in evaluating alternative options, and results in task shedding”, (Kowalski-Trakofler et al.2003, p282).
They mentioned a study that identifies emergency decision makers’ behaviours under stress. This study concluded that these people under stress “not only have the effects of their own stress response and its resulting consequences, the information they must base their judgments on is often unclear, faulty and incomplete”, (Kowalski-Trakofler et al. 2003, p. 289). Bertini (2005) states that, there are three main groups of factors which affect the speed of the vehicles.
The first factor is static, such as the carrying capacity of the roads (e. g. number of road lanes), number of traffic lights and intersections. The second factor dynamic and consists of the number of vehicles on the roads. There is yet another factor, a random which consists of accidents, special events and weather conditions. Borri and Cera (2005) interviewed ambulance drivers in order to seek out the reasons for choosing one route to reach the incident locations instead of another.
The interview results show that the main elements involved in the choice of the routes are width of the roads, presence of street markets, presence of schools, presence of priority lanes and the conditions of the road surface. According to them the width of the roads is important because it makes it easier for the driver to manoeuvre. All of the drivers being interviewed wanted the number of priority lanes to be increased. They also added that the presence of street markets was one of the worst problems of road circulation and near-by roads, because they are mostly used as car parks.
The presence of schools also makes it hard to pass the roads during the hours of entry and exit due to the large nu1mber of people crossing the streets and the number of vehicles on the roads. The Network Analyst extension is used in conjunction with the ArcView 3. 2 desktop GIS package to implement this system.
This was used to find the quickest route or least cost impedance on roads networks based on different variables associated to the roads network from the hospital location to any incident location on road networks. The results were mapped using ArcGIS 9.1, by exporting the analysed results as a shapefile from ArcView 3. 2. For the evaluation procedure, a Dell GPS Navigation System which is embedded within Dell Axim X51v PDA and TomTom GPS, were used. For this research, network model within ArcView consists of a line layer (representing roads) which is made up of segments and nodes), point layer (representing hospitals and incident locations) and rules of the roads. Road networks type classifications includes motorway, road A, road B, minor road, local street, private road – publicly accessible, private road – restricted access, alley and pedestrian street.
While, the nature of road classifications include: Single carriageways, dual carriageways, slip roads, roundabouts and traffic islands. For the purpose of this research, a sample of 3*3 KM of this network was clipped for faster analysis and ease of the field evaluation procedure. A set of roads rules was made prior to the analysis which includes the following. One-way streets that show the flow directions of vehicles along the roads were added to the network. Restricted access streets’ rule was also set to the road networks.
The locations of hospitals were added to the road networks which represent the dispatch point of ambulance vehicles to any incident location on the network. Speed limits in Kilometres were also assigned to each road based upon the nature of the roads (such as dual carriageway, single carriageway, enclosed travel area, roundabouts, slip roads, etc) rather than the type of roads for more accuracy. For example, a road type might have different nature of roads, in which each has a different speed limit.
Based on the speed limits and lengths of roads the travel time cost in minutes was calculated for each road by using this formula (Time = road length / Speed). By setting a set of rules from expert knowledge, it would be possible to find a more accurate travel time along each road, which would consider different criteria acquired from ambulance drivers knowledge affecting the travel (response) time. The system also takes into account the general rules such as the day and time. For example, certain roads are congested at specific times.
As a specific case it is seen that an region of 200 metres around specific schools will be congested during the hours of entry and exit. The system also takes the specific rules into account which affect the speed of an ambulance vehicle such as number of lanes, number of traffic lights, restricted access, presence of schools, street markets, the width of the road and conditions of the road surface. Two terms will be used while integrating the expert knowledge to the system. These are travel time and weighted travel time.
Travel time only considers the length and the speed of each road while weighted travel time considers both the travel time and the rules which have already been set from the experts. Roads and streets within the logical network were given various weights according to the rules mentioned above. A weight is a numerical value that is used along each road. The Weighted roads are then used to calculate the weighted travel time costs which consider all the rules that could affect the speed of the vehicles in reality.
For example, if a road has one lane, it will be given more weight than the two lanes road. The travel costs of these roads are then recalculated by multiplying the weight with the travel cost. As a result of this procedure, the calculated travel time will be higher in one lane road than two lane road; i. e. one lane roads will be inadvisable because a vehicle is expected to take longer time to travel along these roads than the two lanes and will have an effect on manoeuvre ability of ambulance vehicles along these roads.
This procedure is then repeated for all other rules in order to find the travel time costs for the whole network (table 2). Note: For the analysis purpose of this research, travel time field’s name and weighted travel time fields’ names were changed into MINUTES, INCHES, FEET and MILES, in order to be identified as a cost field by Network Analyst. Finally, the general rules which consider the day of the week and time of day are integrated. This is done by adding other weights to the roads, dependant on the day and the time, and after this the real travel time cost will be updated.