Department of Smart Mobility
Indian Institute of Technology Hyderabad
Spot Speed Study Assignment
Submitted By
Rishabh Chaturvedi - SM24MTECH11008
Traffic Engineering and Intelligent Transportation Systems
SM5043
April 20, 2025
1
1 Problem Statement
This study conducts a comparative analysis of vehicular speed data obtained through
two distinct measurement methodologies: field-based RADAR/LiDAR and TIRTL (The
Infra-Red Traffic Logger) sensor systems. The investigation seeks to analyze speed dis-
tributions and calculate specific percentile speeds (15th, 50th, 85th, and 98th) that are
critical for traffic engineering applications.
2 Objectives
The objectives of this spot speed study are:
1. To develop comprehensive frequency and cumulative frequency distributions for
speed data collected via both methodologies
2. To determine and compare key percentile speeds across both measurement systems
3. To evaluate the relative accuracy, reliability, and potential limitations of each data
collection method
4. To assess the implications of any observed differences for traffic management and
road safety applications
Through statistical analysis of the speed data, this study aims to contribute to the
understanding of measurement system efficacy in traffic monitoring contexts. The find-
ings may inform best practices in data collection methodology selection for transportation
engineering applications where precise speed percentile calculations are essential.
3 Methodology
This study employs a quantitative approach to analyze and compare vehicular speed data
collected through two different measurement technologies. The methodology encompasses
data collection procedures, analytical techniques, and comparative assessment methods.
3.1 Data Collection
The speed data analyzed in this study was collected using two distinct methodologies:
• Field Measurement (RADAR/LiDAR): A total of 140 speed observations
were collected manually using a RADAR/LiDAR based speed gun. This device
measures vehicle speeds by detecting the Doppler shift in the reflected signal or by
calculating the time difference between multiple laser pulses.
• TIRTL (The Infra-Red Traffic Logger): A comprehensive dataset of 23,898
spot speed measurements collected from TIRTL was provided for analysis. This
system utilizes infrared beams positioned across the roadway to detect vehicle pres-
ence, speed, and classification. TIRTL operates by measuring the time taken for a
vehicle to break sequential infrared beams, thereby calculating speed based on the
known distance between beams.
2
Data was collected at the same location and during comparable traffic conditions to
ensure valid comparison between the two methodologies. The analysis focused exclusively
on the speed data column from the provided Excel sheets, as specified in the assignment
guidelines.
3.2 Analytical Approach
The following analytical procedures were applied to the data from both measurement
methods:
1. Frequency Distribution Analysis: The raw speed data was organized into ap-
propriate speed bins to create frequency distribution tables. This involved calcu-
lating the number and percentage of vehicles within each speed range.
2. Graphical Representation: Frequency distribution curves were plotted to visual-
ize the distribution pattern of vehicle speeds for both measurement methodologies.
Similarly, cumulative frequency distribution curves were created to facilitate per-
centile speed calculations.
3. Percentile Speed Calculation: The 15th, 50th, 85th, and 98th percentile speeds
were determined using linear interpolation on the cumulative frequency distribu-
tions. These percentiles were selected as they represent standard reference points
in traffic engineering:
• 15th percentile: Represents the lower speed threshold
• 50th percentile: Represents the median speed (average driver behavior)
• 85th percentile: Commonly used for setting speed limits
• 98th percentile: Represents high-end speeding behavior
4. Comparative Analysis: Results from both methodologies were systematically
compared to identify differences in distribution patterns and percentile speeds. Sta-
tistical measures were employed to quantify the degree of variance between the two
methods.
3.3 Analytical Tools
The analysis was conducted using standard statistical techniques implemented in Mi-
crosoft Excel. Linear interpolation was employed for precise percentile determination
using the formula:
(y − y1 )
x = x1 + × (x2 − x1 ) (1)
(y2 − y1 )
Where:
• x is the speed corresponding to the desired percentile
• y is the percentile of interest (15, 50, 85, or 98)
• (x1 , y1 ) and (x2 , y2 ) are the known points from the cumulative frequency distribution
that bracket the desired percentile
3
4 Data Analysis Of TIRTL Sensor Method
4.1 Frequency Distribution Table
Speed Range Mid Speed Frequency Frequency Cumulative
(Kmph) (Kmph) (f) (%) Frequency (%)
0-5 2.5 2 0.008 0.008
5-10 7.5 24 0.100 0.109
10-15 12.5 98 0.410 0.519
15-20 17.5 136 0.569 1.088
20-25 22.5 197 0.824 1.912
25-30 27.5 461 1.929 3.841
30-35 32.5 1085 4.540 8.381
35-40 37.5 1843 7.712 16.093
40-45 42.5 2896 12.118 28.212
45-50 47.5 3314 13.867 42.079
50-55 52.5 3258 13.633 55.712
55-60 57.5 2994 12.528 68.240
60-65 62.5 2572 10.762 79.002
65-70 67.5 1793 7.503 86.505
70-75 72.5 1322 5.532 92.037
75-80 77.5 905 3.787 95.824
80-85 82.5 518 2.168 97.991
85-90 87.5 240 1.004 98.996
90-95 92.5 125 0.523 99.519
95-100 97.5 50 0.209 99.728
100-105 102.5 43 0.180 99.908
105-110 107.5 13 0.054 99.962
110-115 112.5 7 0.029 99.992
115-120 117.5 1 0.004 99.996
120-125 122.5 0 0.000 99.996
125-130 127.5 1 0.004 100.000
Table 1: Frequency Distribution of Spot Speed Data Using TIRTL
4.2 Analysis of TIRTL Spot Speed Data
Table 1 shows the distribution of vehicle speeds measured using the TIRTL system. This
data includes 23,898 vehicle speed readings, grouped into 5 km/h speed ranges from 0 to
130 km/h.
Most vehicles travel between 40 and 65 km/h. The most common speed range is 45-50
km/h, with 3,314 vehicles (about 14% of all vehicles). The table shows that about 56%
of vehicles travel at speeds of 52.5 km/h or below, and about 80% travel at or below 62.5
km/h.
Very few vehicles drive at extremely high speeds - only about 2% of vehicles go faster
than 85 km/h. Similarly, very few vehicles (less than 1%) travel slower than 15 km/h,
which suggests traffic was flowing well during the measurement period.
4
4.3 Frequency Distribution Analysis
Figure 1: Frequency Distribution Curve of Vehicle Speeds Measured by TIRTL
Figure 2: Cumulative Frequency Distribution Curve Measured by TIRTL
5
Table 2: Key Percentile Speeds from TIRTL Measurements
Percentile Speed (km/h)
15th 36.79
50th 50.41
85th 66.50
98th 82.55
4.4 Statistical Analysis of Vehicle Speeds
4.4.1 Measures of Central Tendency
Based on the frequency distribution data of the TIRTL measurements, the following
statistical parameters were calculated:
• Mean Speed: The average speed was calculated using the weighted average for-
mula:
(xi · fi )
P
x̄ = P (2)
fi
Where xi represents the mid-speed value and fi represents the frequency.
x̄ = 1,212,233.5
23,898
= 50.72 km/h
• Median Speed: The 50th percentile speed, calculated through linear interpola-
tion, is 50.41 km/h. This represents the middle value in the dataset, with half of
the vehicles traveling faster and half traveling slower.
• Modal Speed: The speed range with the highest frequency is 45-50 km/h (with
3,314 observations), giving a mid-speed value of 47.5 km/h. This represents the
most commonly observed vehicle speed.
4.4.2 Key Percentile Speeds
The key percentile speeds derived from the cumulative frequency distribution are pre-
sented in Table 2.
These percentile values have significant implications for traffic engineering applica-
tions:
• The 15th percentile speed (36.79 km/h) represents slower-moving vehicles and may
indicate the lower boundary of the speed range that most drivers consider reason-
able.
• The 50th percentile speed (50.41 km/h) indicates the median driving behavior under
the existing roadway conditions.
• The 85th percentile speed (66.50 km/h) is particularly important as it is often used
as a reference for setting appropriate speed limits. This value represents the speed
that the majority of drivers consider reasonable and safe given the roadway design
and conditions.
6
• The 98th percentile speed (82.55 km/h) identifies the upper boundary of typical
driving behavior, with only 2% of vehicles exceeding this speed. These vehicles may
represent aggressive driving behavior or special circumstances.
The difference between the 85th and 15th percentile speeds (66.50 - 36.79 = 29.71
km/h) provides insight into the speed variance on this roadway. This relatively large
range suggests moderate speed diversity, which may have implications for traffic flow and
safety.