Transportation Systems Engineering

6. Measurement over a Short Section

Chapter 6 Measurement over a Short Section

The main purpose of this chapter is to determine trac parameter, specially speed. Speed measurements are most often taken at a point (or a short section) of road way under conditions of free ow. The intent is to determine the speeds that drivers select, unaected by the existence of congestion. This information is used to determine general speed trends, to help determine reasonable speed limits, and to assess safety.

6.2

Speed Studies

As speed denes the distance travelled by user in a given time, and this is a vibrant in every trac movement. In other words speed of movement is the ratio of distance travelled to time of travel. The actual speed of trac ow over a given route may uctuated widely, as because at each time the volume of trac varies. Accordingly, speeds are generally classied into three main categories 1. Spot speed: This is the instantaneous speed of a vehicle at any specic location. 2. Running speed: This is the average speed maintained over a particular course while the vehicle is in the motion. 3. Journey speed: This is the eective speed of the vehicle on a journey between two points and the distance between two points and the distance between these points divided by the total time taken for the vehicle to complete the journey, it includes all delay.

Dr. Tom V. Mathew, IIT Bombay

6.1

February 9, 2013

Transportation Systems Engineering Trac Stream Average Speed (in Kmph) below 15 15 -25 above 25

6. Measurement over a Short Section Recommended Study Length (in m) 30 60 90

6.3

Spot Speed Studies

When we measure the trac parameter over a short distance, we generally measure the spot speed. A spot speed is made by measuring the individual speeds of a sample of the vehicle passing a given spot on a street or highway. Spot speed studies are used to determine the speed distribution of a trac stream at a specic location. The data gathered in spot speed studies are used to determine vehicle speed percentiles, which are useful in making many speed-related decisions. Spot speed data have a number of safety applications, including the following 1. Speed trends, 2. Trac control planning, 3. Accidental analysis, 4. Geometric design, 5. Research studies.

6.4

Methods of Measurement

Methods of conducting spot speed Studies are divided into two main categories: 1. Manual 2. Automatic Spot speeds may be estimated by manually measuring the time it takes a vehicle to travel between two dened points on the roadway a known distance apart (short distance), usually less than 90m. Distance between two points is generally depending upon the average speed of trac stream. Following tables gives specic distance for particular speed range. Following are the some methods to measure spot speed of vehicles in a trac stream, in which rst two are manual methods and other are automatic: Dr. Tom V. Mathew, IIT Bombay 6.2 February 9, 2013

Transportation Systems Engineering

6. Measurement over a Short Section

6.4.1

Pavement markings:

In this method, markings of pavement are placed across the road at each end of trap. Observer start and stops the watch as vehicle passes lines. In this method, minimum two observers required to collect the data, of which one is stand at the starting point to start and stop the stop watch and other one is stand at end point to give indication to stop the watch when vehicle passes the end line. Advantages of this method are that after the initial installation no set-up time is required, markings are easily renewed, and disadvantage of this is that substantial error can be introduced, and magnitude of error may change for substitute studies and this method is only applicable for low trac conditions.
Vertical Reference point End Timing Vertical Reference point

Study length Approaching Vehicle

Start timing Observer 2 X Observer 1

Figure 6:1: Pavement Marking

6.4.2

Enoscope or Mirror box:

Enoscope consists of a simple open housing containing a mirror mounted on a tripod at the side of the road in such a way that an observers line of sight turned through 90o. The observer stands at one end of section and on the other end enoscope is placed and measure the time taken by the vehicle to cross the section (g 6.2). Advantages of this method are that it simple and eliminate the errors due to parallax and considerable time is required to time each vehicle, which lengthen the study period and under heavy trac condition it may be dicult to relate ostentatious to proper vehicle are the disadvantages of enoscope method.

6.4.3

Road Detector (Pressure contact strips)

Pressure contact strips, either pneumatic or electric, can be used to avoid error due to parallax and due to manually starting and stopping the chronometer or stopwatch. This is the best

Dr. Tom V. Mathew, IIT Bombay

6.3

February 9, 2013

Transportation Systems Engineering

6. Measurement over a Short Section

observer

Base length

Figure 6:2: Enoscope Method method over short distance it gives quite relevant data and if it is connected through graphical recorder then it gives continuous data automatically.

6.4.4

Doppler-Principle Meters (Radar)

This is recently developed method, it automatically records speed, employs a radar transmitterreceiver unit. The apparatus transmits high frequency electromagnetic waves in a narrow beam towards the moving vehicle, and reected waves changed their length depending up on the vehicles speed and returned to the receiving unit, through calibration gives directly spot speed of the vehicle.

6.4.5

Electronic-Principle Detectors (Photography):

In this method a camera records the distance moved by a vehicle in a selected short time. In this exposure of photograph should be in a constant time interval and the distance travelled by the vehicle is measured by projecting the lms during the exposure interval. The main advantage of method that, it gives a permanent record with 100% sample obtained. This method is quite expensive and generally used in developed cities. In this we can use video recorder which give more accurate result.

6.5

Data Collection Sheets

The measured data by the above techniques should be collected into some formats, following are the some types of data collection sheets which are used for manual and automatic methods, 1. For Enoscope and Pavement Marking Methods Dr. Tom V. Mathew, IIT Bombay 6.4 February 9, 2013

Transportation Systems Engineering

6. Measurement over a Short Section

S NO. 1 2 3 4 5 6 7 8 9 10

TIME TAKEN (in sec) Car/jeep 3 Wheeler 2 Wheeler Cycle LCV HCV

Surveyor:

Figure 6:3: Data collection sheet for Enoscope and Pavement Marking Methods 2. For automatic methods
Spot Speed Data Collection Form Location: Weather: Type of road: Data: Time: Measurement Technique:

Vehicle No.

Speed

Vehicle No.

Speed

Surveyor:
Spot Speed Data Collection Form Location: Weather: Type of road: Measurement Technique: Date: Time: Base Length:

Speed Range in kmph 0 5 10 15 20 25 30 35 40 45 50 55 60 5 10 15 20 25 30 35 40 45 50 55 60 65

Tally Mark Car/Jeep Bus/Truck Car/Jeep

Number Bus/Truck

Total

Surveyor:

Figure 6:4: Data collection sheet for Automatic Methods

Dr. Tom V. Mathew, IIT Bombay

6.5

February 9, 2013

Transportation Systems Engineering

6. Measurement over a Short Section

6.6

Data Presentation

From the above methods, the collected data have to present into the some representable form, this makes its calculation and analysis simpler and easier. The following methods to present the spot speed data:

6.6.1

Frequency Distribution Table

After the collection of data in the given conditions, arrange the spot speed values in order to their magnitudes. Then select an interval speed (e.g. 5 Kmph) and make grouping of data which come under this range. Now, prepare the frequency distribution table.

6.6.2

Frequency Distribution Curve

For each speed group, the % frequency of observations within the group is plotted versus the middle (midmark) speed of the group(s). As shown in Fig 6.5. From this curve the modal speed and pace of trac ow can be determine. Generally the shape of the curve follows the normal distribution curve, this because the most of the vehicles move on road near by mean speed and very few deviate from mean speed.

6.6.3

Cumulative Frequency Distribution Curve

For each speed group, the % cumulative frequency of observations is plotted versus the higher limit of the speed group (Fig 6.5). The cumulative frequency distribution curve, however, results in a very useful plot of speed versus the percent of vehicles traveling at or below the designated speed. For this reason, the upper limit of the speed group is used as the plotting point. In both the distribution curve, the plots are connected by a smooth curve that minimizes the total distance of points falling above the line and those falling below the line. A smooth curve is dened as one without.

6.7

Distribution Characteristics

Common descriptive statistics may be computed from the data in the frequency distribution table or determined graphically from the frequency and cumulative frequency distribution curves. These statistics are used to describe two important characteristics of the distribution:

Dr. Tom V. Mathew, IIT Bombay

6.6

February 9, 2013

Transportation Systems Engineering

6. Measurement over a Short Section

6.7.1

Measure of Central Tendency

Measure which helps to describe the approximate middle or centre of the distribution. Measures of central tendency include the average or mean speed, the median speed, the modal speed, and the pace.
25 Mode $\% Frequency 20 15 10 5 0 100 $Cum.\%freq$ 90 80 70 60 50 40 30 20 10 0 $86\%$ $\% Veh in pace =8614= 72\%$ Median Pace

32

36

40

44

48

52

56

60

64

Speed (kph)

$14\%$ Speed (kph)

32

36

40

44

48

52

56

60

64

Figure 6:5: Frequency and Cumulative Frequency Distribution curve

The Mean Speed The arithmetic (or harmonic) average speed is the most frequently used speed statistics. It is the measure of central tendency of the data. Mean calculated gives two kinds of mean speeds. vm = fi vi n (6.1)

Where, vm is the mean or average speed, vi is the individual speed of the ith vehicle, fi is the frequency of speed, and n is the total no of vehicle observed (sample size). Time mean Speed: If data collected at a point over a period of time, e.g. by radar meter or stopwatch, produce speed distribution over time, so the mean of speed is time mean speed. Space mean Speed: If data obtained over a stretch (section) of road almost instantaneously, aerial photography or enoscope, result in speed distribution in space and mean is space mean speed. Distribution over space and time are not same. Time mean speed is higher than the space mean speed. The spot speed sample at one end taken over a nite period of time will tend to include some fast vehicles which had not yet entered the section at the start of the survey, but will exclude some of the slower vehicles. The relationship between the two mean speeds is expressed by: 2 vtm = vsm + (s /vsm ) (6.2) Dr. Tom V. Mathew, IIT Bombay 6.7 February 9, 2013

Transportation Systems Engineering

6. Measurement over a Short Section

where, vtm and vsm are the time mean speed and space mean speed respectively. And s is the standard deviation of distribution space. The Median Speed The median speed is dened as the speed that divides the distribution in to equal parts (i.e., there are as many observations of speeds higher than the median as there are lower than the median). It is a positional value and is not aected by the absolute value of extreme observations. By denition, the median equally divides the distribution. Therefore, 50% of all observed speeds should be less than th e median. In the cumulative frequency curve, the 50th percentile speed is the median of the speed distribution. Median Speed = v50 The Pace The pace is a trac engineering measure not commonly used for other statistical analyses. It is dened as the 10Km/h increment in speed in which the highest percentage of drivers is observed. It is also found graphically using the frequency distribution curve. As shown in g 6.5. The pace is found as follows: A 10 Km/h template is scaled from the horizontal axis. Keeping this template horizontal, place an end on the lower left side of the curve and move slowly along the curve. When the right side of the template intersects the right side of the curve, the pace has been located. This procedure identies the 10 Km/h increments that intersect the peak of the curve; this contains the most area and, therefore, the highest percentage of vehicles. The Modal Speed The mode is dened as the single value of speed that is most likely to occur. As no discrete values were recorded, the modal speed is also determined graphically from the frequency distribution curve. A vertical line is dropped from the peak of the curve, with the result found on the horizontal axis.

6.7.2

Measure of Dispersion

Measures describe the extent to which data spreads around the centre of the distribution. Measures of dispersion include the dierent percentile speeds i.e. 15th, 85th,etc. and the standard deviation.

Dr. Tom V. Mathew, IIT Bombay

6.8

February 9, 2013

Transportation Systems Engineering Standard Deviation:

6. Measurement over a Short Section

The most common statistical measure of dispersion in a distribution is the standard deviation. It is a measure of how far data spreads around the mean value. In simple terms, the standard deviation is the average value of the dierence between individual observations and the average value of those observations. The Standard deviation, s , of the sample can be calculated by crs = The 85th and 15th Percentile Speeds The 85th and 15th percentile speeds give a general description of the high and low speeds observed by most reasonable drivers. It is generally thought that the upper and lower 15% of the distribution represents speeds that are either too fast or too slow for existing conditions. These values are found graphically from the cumulative frequency distribution curve of Figure 6.4. The curve is entered on the vertical axis at values of 85% and 15%. The respective speeds are found on the horizontal axis. The 85th and 15th percentile speeds can be used to roughly estimate the standard deviation of the distribution est , although this is not recommended when the data is available for a precise determination. v8 5 v1 5 (6.4) est = 2 The 85th and 15th percentile speeds give insight to both the central tendency and dispersion of the distribution. As these values get closer to the mean, less dispersion exists and the stronger the central tendency of the distribution becomes. The 98th Percentile Speed The 98th Percentile Speed is also determining from the cumulative frequency curve, this speed is generally used for geometric design of the road. fi (vi vv )2 n1 (6.3)

6.8
6.8.1

Data Analysis
Standard Error of the mean

The means of dierent sample taken from the same population are distributed normally about the true mean of population with a standard deviation, is known as standard error. s (6.5) Se = n Dr. Tom V. Mathew, IIT Bombay 6.9 February 9, 2013

Transportation Systems Engineering

6. Measurement over a Short Section

6.8.2

Sample Size:

Generally, sample sizes of 50 to 200 vehicles are taken. In that case, standard error of mean is usually under the acceptable limit. If precision is prior then minimum no. of sample should be taken, that can be measured by using the following equation. nr =
2 Z 2 s 2 Se

(6.6)

where, nr is the no. of sample required, s is the Standard deviation, Z is value calculated from Standard Normal distribution Table for a particular condence level (i.e. for 95% condence Z=1.96 and for 99.7% condence Z=3.0) and Se is the permissible (acceptable) error in mean calculation.

6.8.3

Precision and Condence Intervals

Condence intervals express the range within which a result for the whole population would occur for a particular proportion of times an experiment or test was repeated among a sample of the population. Condence interval is a standard way of articulate the statistical accuracy of an experiment based assessment. If assess has a high error level, the equivalent condence interval will be ample, and the less condence we can have that the experiment results depict the situation among the whole population. When quoting condence It is common to refer to the some condence interval around an experiment assessment or test result. So, the condence interval for estimated true mean speed can be calculated by = vm Zs (6.7)

where, is the condence interval, vm is mean speed, s is standard deviation and Z is constant for specied condence.

6.9

Location for Speed Studies

The speed studies are accompanied for eminently logical purposes that will inuence what trac engineering measures are implemented in any given case. The location at which speed measurements are taken must conform to the intentional purpose of the study. The guiding philosophy behind spot speed studies is that measurements should include drivers freely selecting their speeds, unaected by trac congestion. For example if driver approaches to a toll plaza, then he has to slow his speed, so this is not suitable location to conduct the study, measurements should be taken at a point before drivers start to decelerate. Similarly, if excessive speed Dr. Tom V. Mathew, IIT Bombay 6.10 February 9, 2013

Transportation Systems Engineering

6. Measurement over a Short Section

around a curve is thought to be contributing to o-the-road accidents, speed measurements should be taken in advance of the curve, before deceleration begins. It may also be appropriate, however, to measure speeds at the point where accidents are occurring for evaluation with approach speeds. This would allow the trac engineer to assess whether the problem is excessive approach speed or that drivers are not decelerating suciently through the subject geometric element, or a combination of both. A study of intersection approach speeds must also be taken at a point before drivers begin to decelerate. This may be a moving point, given that queues get shorter and longer at dierent periods of the day.

6.10

Example Problem

Example1 Consider the following spot speed data, collected from a freeway site operating under free-ow conditions: 1. Plot the frequency and cumulative frequency curves for these data. 2. Find and identify on the curves: median speed, modal speed, pace, percent vehicles in pace. 3. Compute the mean and standard deviation of the speed distribution. 4. What are the condence bounds on the estimate of the true mean speed of the underlying distribution with 95% condence? With 99.7% condence? 5. Based on the results of this study, a second is to be conducted to achieve a tolerance of 1.5 km/h with 95% condence. What sample size is needed? Solution: For the spot speed study, rst draw a frequency distribution table (Table 6.3). 1. From the table 6.3, we can draw frequency distribution and cumulative frequency distribution curve.(shown in Fig 6.6 and 6.7)

2. From the curves, Median speed, v5 0 = 43 Kmph Modal speed, = 38 Kmph Pace = 33 - 43 Kmph Percent vehicles in pace = 54-20= 34% 85th Percentile speed = 58 Kmph Dr. Tom V. Mathew, IIT Bombay 6.11 February 9, 2013

Transportation Systems Engineering

6. Measurement over a Short Section

Speed Range Frequency 21 25 2 26 30 6 31 35 18 36 40 25 41 45 19 46 50 16 51 55 17 56 60 12 61 65 7 66 70 4 71 75 3 76 80 1

Speed Range Mid speed Vi 21 25 23 26 30 28 31 35 33 36 40 38 41 45 43 46 50 48 51 55 53 56 60 58 61 65 63 66 70 68 71 75 73 76 80 78 Total

Frequency fi 2 6 18 25 19 16 17 12 7 4 3 1 130

% frequency 2% 5% 14% 19% 15% 12% 13% 9% 5% 3% 2% 1% 100%

Cumulative % Frequency 2% 6% 20% 39% 54% 66% 79% 88% 94% 97% 99% 100%

fi Vi 46 168 594 950 817 768 901 696 441 272 219 78 5950

fi (Vi 1036 1894 2934 1509 145. 79.6 888. 1795 2078 1976 2224 1038 1760

Table 6:1: Solution of the example problem

Dr. Tom V. Mathew, IIT Bombay

6.12

February 9, 2013

Transportation Systems Engineering

6. Measurement over a Short Section

20\% Frequency(\%)

Mode pace

15\%

10\%

5\%

0\% 0 10 20 30 33

38 43 40

50

60

70 80 Speed (kmph)

90

Figure 6:6: Frequency Distribution Curve

$v_85$

$v_15$

43
50

58
60 70

Speed (kmph)

80

90

Figure 6:7: Cumulative Frequency Distribution Curve 3. Mean is calculated by using vm = vm vm Standard Deviation of the Speed s = s fi (vi vm )2 n1 17603.08 = 130 1 = 11.7Kmph fi vi n 5950 = 130 = 45.77Kmph

4. The condence bounds on the estimate of the true mean speed of the underlying distribution are: = vm Zs (6.8) Dr. Tom V. Mathew, IIT Bombay 6.13 February 9, 2013

Transportation Systems Engineering Parameter Median speed Modal speed Pace vehicles in pace Mean speed Standard Deviation 85th percentile 15th percentile Speed 98th percentile Speed Condence interval For 95% . For 99.7% Required sample Size

6. Measurement over a Short Section Value 43 Kmph 38 Kmph 33-43 Kmph 34% 45.77 Kmph 11.7 Kmph 58 Kmph 32 Kmph 72 Kmph 45.7722.93 Kmph 45.7735.1 Kmph 234

Table 6:2: Result of the example problem (a) For 95% condence, Z= 1.96 = 45.77 1.96 11.7 = 45.77 22.93kmph (b) For 99.7% condence, Z= 3.0 = 45.77 3.0 11.7 = 45.77 35.1kmph 5. Sample size required for 95% condence with acceptable error of 1.5 Kmph nr = nr nr
2 Z 2 s 2 Se 1.962 11.72 = 1.52 = 234

So, given sample size is not sucient and we require minimum 234 samples to achieve that condence with given acceptable error. The results are summaries in table xxx

Dr. Tom V. Mathew, IIT Bombay

6.14

February 9, 2013

Transportation Systems Engineering

6. Measurement over a Short Section

6.11

Summary

This chapter has presented the basic concepts of speed studies. Spot speed studies are conducted to estimate the distribution of speeds of vehicle in the trac stream at a particular position on highway. This is done by recording the speeds of vehicle at the specied location. These data are used to obtain speed characteristics such as mean speed, modal speed, pace, standard deviation and dierent percentile of speeds. The important factors which should consider during plan of studies is the location of study, time and duration of study. The data sample collected should contain samples size. These gives precision and accuracy of result.

6.12

References

1. F D Hobbs. Trac Planning and Engineering. Pergamon Press, 1979. 2nd Edition. 2. Nicholas J Garber Lester A Hoe. Trac and Highway Engineering. Cengage Learning Product, Fourth Edition, 2009. 3. Theodore M Matson, Wilbure S smith, and Fredric W Hurd. Trac engineering, 1955. 4. R P Roess, S E Prassas, and W R McShane. Trac Engineering. Pearson Education International, 2005.

Dr. Tom V. Mathew, IIT Bombay

6.15

February 9, 2013