Excepciones en caso de urgencia

The aim of this subsection is to reveal the relation between the second vehicle delay phenomenon and the traditional start-up lost time problem in the queuing vehicle discharge process.

4.2.1 Video record survey of queue discharge

In Section 1.3 research problem introduction, evidence for the second vehicle delay only uses data from Brisbane. To offer more visual evidence for the second vehicle delay phenomenon in queue discharge, further video recording of the queuing vehicle departures at urban signalised intersections was conducted in Brisbane, Melbourne and Sydney (Australia).

Table 6: Video record survey of the second vehicle delay phenomenon

Video clips Record date & place Aim

21 August 2008, Vulture Street and Grey Street, Southbank, Brisbane. Recorded by Shuai Yang

Different time but same location

16 November 2010, Vulture Street and Grey Street, Southbank, Brisbane. Recorded by Shuai Yang

Different time but same location

20 September 2008, Nicholson Street and Victoria Street, Melbourne. Recorded by Shuai Yang

In the same city and same day but different location

20 September 2008, Brunswick Street and Victoria Parade, Fitzroy, Melbourne. Recorded by Shuai Yang

In the same city and same day but at different location

26 July 2010, Broadway and Abercrombie Street, Ultimo, Sydney. Recorded by Haoyao Zhang

Different time and different location

To ensure the accuracy of this survey, the video recording is based on the following three criteria: different time and different location, different time but same location, in the same city on the same day but different location. To avoid affecting drivers, this survey used a mobile phone to record the video clips which only captured the vehicle discharge process during one signal phase when the researcher arrived at the intersection. Some clips are recorded in Melbourne in 2008, and the Sydney video is recorded by a friend Haoyao Zhang. For future consideration, one limitation of this research is that it only focuses on exploring the second vehicle delay phenomenon in Australia. Hence, it is proposes to widely conduct this research in other countries in the future.

Table 6 indicates that the second vehicle in the queue generally starts to accelerate a few seconds after the first vehicle, while the third vehicle accelerates almost at the same time as the second. Therefore, the second driver response time is longer than that of other drivers. In the Brisbane case, the problem still exists, even two years on.

The video in the Sydney case indicates that the second taxi driver, a professional driver, starts to accelerate only when the first vehicle has crossed the stop line. However, the survey only offers visual evidence, so it is necessary to conduct quantitative measurements and analysis to support this research. The research problems of this study are again strengthened.

4.2.2 Data analysis of second vehicle delay

Traditional methods use headway to measure queuing vehicle dynamic performance at signalised intersections (J A Bonneson, 1992; Fairclough et al., 1997; Jin et al., 2009; Lee & Chen, 1986). Greenshield et al. (1947) stated that driver response time to signal change and response time between successive vehicles should be calculated independently in discharge process research. Therefore, this research explores each queuing vehicle’s dynamic performance in the discharge process by considering two aspects: the time it takes the driver to perceive a signal change (or obtain safety distance to accelerate) and react, and the time it takes the vehicle to start to move and crosses the stop line.

Figure 27: Relationship zi & ki with hi

These two components are separately symbolised as zi and ki in this thesis. This

research also uses hi to represent the time for the ith vehicle to cross the stop line

added to ki equals the time hi. The headway at the stop line will be hi minus hi-1.

Figure 27 presents the relationship between zi, ki and hi (Yang & Chung, 2012).

To characterise the leading vehicles’ dynamic performance in the discharge process, this stage uses the obtained Peachtree Data five leading vehicles’ dynamic parameters to explore the second vehicle delay phenomenon. Typically, most departure models use average methods to obtain driver response time, mainly because the variation of driver response time is presumed to be small. Thus, the definition of the term “Stop” (less than 5 km/h) criterion is used to extract the driver response time. Table 7 presents the five leading queuing vehicles’ average response time zi and the headway. This stage uses average times to measure the time that the

vehicle spends in moving from queuing position to the stop line (the time ki). Using

average time for this research can cover most common samples because, with the exception of the first vehicle, each queuing vehicle needs to have a safety distance before it can accelerate.

Table 7: Peachtree Street data set yield time zi & ki

Vehicle position Response time zi (Second)

Time to cross the stop line ki

(Second)

Total time (Second)

Headway (Second)

The first five queued vehicles at signalised intersections

The time the vehicle starts to move from green onset

The time the vehicle takes to cross the stop line from queue position

The total time the vehicle takes to cross the stop line from green onset The headway between each vehicle 1st vehicle 1.5 0 1.5 1.5 2nd vehicle 2.7 2.0 4.7 3.2 3rd vehicle 3.3 4.2 7.5 2.8 4th vehicle 4.3 5.5 9.8 2.3 5th vehicle 6 6.3 12.3 2.5

In this step, the Peachtree Data study reveals that the time zi and the average time ki

are regressed to generate the headway time. The waving trend in driver response time (zi) in the above analyses reveals that the driver response time is not a constant

parameter, and a significant headway exists between the first and the second vehicle. This fits the observed visual display (see Section 4.2.1, Table 6).