SlideShare a Scribd company logo
Origin Destination Surveys OD调查
…and Trip Generation Surveys…
交通出行率调查
Project
Description 项
目概况
Project Setting
项目区域现状
Travel Demand
Analysis
交通需求分析
Transportation
Impact Analysis
交通影响分析
Transportation Mitigation Measures
交通改善措施
Appendix A, Appendix B, etc.
etc. 附录A, 附录B, 等等
But first…但首先,
Traffic Counts, REVISITED…
高峰时段交叉口交通流量调查,回访
Scoping Site in Advance for
Intersection Traffic Counts
计算前的准备工作
 Visit the project site
 take pictures of traffic controls and lanes 拍摄交通控制
和路面情况
 observe intersection conditions
 Minimize Complications in taking traffic counts
 Engineer should make determination of staff needs, in
advance.
 Timing of intervals, Recording
 Complications means LOST DATA and a bad count on the
day of survey.
 Incorrect Counts
 Missing Information.Wrong North Arrow, etc
Traffic Count Sheet: Hard to Interpret
交通流量调查问卷: 用中文很难解释
Problem:
Intersection mis-
labeled because road
closure was at
LongShouLu
Traffic Count Sheet: Incorrect Method
交通流量调查问卷: 错误的方法
Counted first...
5:30-5:45 pm
Counted second...
5:45-6:00 pm
Counted third...
6:00-6:15 pm
Traffic Count Sheet: Hard to Interpret
交通流量调查问卷:用中文 很难解释
Traffic Count Sheet: Only 15 Minute Count
交通流量调查问卷: 只有15分钟的流量调查
15 Minute Count is Not Enough
15分钟流量调查不够
Left
Out
Right
Out
Left In Right In
3:30 12 8 16 12
3:45 20 16 24 28
4:00 20 4 8 24
4:15 32 28 12 24
4:30 32 8 20 48
4:45 16 16 24 52
5:00 28 12 32 64
5:15 32 20 32 88
5:30 40 24 28 92
5:45 16 12 72 104
6:00 40 12 60 116
6:15 32 8 20 60
What you get when you
multiply a 15 minute count by
4:
DIFFERENT RESULTS!
不同的结果
32 17 48 100
32 17 48 100
32 17 48 100
32 17 48 100
32 17 48 100
32 17 48 100
32 17 48 100
32 17 48 100
32 17 48 100
32 17 48 100
32 17 48 100
Actual:
Actual:
Actual:
Actual:
Actual:
Actual:
Actual:
Actual:
Actual:
Actual:
Actual:
15 Minute Count is Not Enough
15分钟流量调查不够
 Long Hu Xi Juan 3 hr. Count. 龙湖西苑3小时流量调查 A real count.
 What happens when you multiply a 15 minute count
by 4 to “short-cut” the count process. 4当你计算15分
钟的流量,会发生什么?You get vastly different
hourly totals!你会得到不同的结果 Likely Incorrect.
很有可能是不正确的
0
20
40
60
80
100
120
140
3:30 3:45 4:00 4:15 4:30 4:45 5:00 5:15 5:30 5:45 6:00 6:15 Actual
Left Out
Right Out
Left In
Right In
Actual Peak HourTotal实际高峰时段总
Actual Peak HourTotal 实际高峰时段总
Actual Peak HourTotal实际高峰时段总
PeakHourVolume
左出口
右出口
左入口
右入口
We need to Count for:
 TWO HOURS minimum
 We find the peak HOUR from
these two hours of data.
 It cannot be guessed…
 …it is incorrect to multiply a 15
minute count by four(4) to get an
hourly total. 15min x 4 peak hour
15 Minute Count is Not Enough
 This example shows that you get four very different answers
 The first 15 minutes x 4 is 15% of actual.
 The second 15 minutes x 4 is 25% of actual.
 The third 15 minutes x 4 is 50% of actual, etc. etc.
 The fourth set was close, but that was LUCKY
0
20
40
60
80
100
120
140
3:30 3:45 4:00 4:15 4:30 4:45 5:00 5:15 5:30 5:45 6:00 6:15 Actual
Left Out
Right Out
Left In
Right In
Actual Peak HourTotal
Actual Peak HourTotal
Actual Peak HourTotal
PeakHourVolume
Photo Documentation is very
Important and Helpful
龙寿路 /人兴路
LongShou / RenXing
Photo Location 1: 1
龙寿路 /人兴路
LongShou / RenXing
1
Photo Location 1: 2
龙寿路 /人兴路
LongShou / RenXing
1
Photo Location 1: 3
龙寿路 /人兴路
LongShou / RenXing
1
Photo Location 2: 1
龙寿路 /人兴路
LongShou / RenXing
2
Photo Location 2: 2
龙寿路 /人兴路
LongShou / RenXing
2
Photo Location 3: 1
龙寿路 /人兴路
LongShou / RenXing
3
Photo Location 3: 2
龙寿路 /人兴路
LongShou / RenXing
3
Photo Location 4: 1
龙寿路 /人兴路
LongShou / RenXing
4
ORIGIN DESTINATION SURVEYS
An Overview of
TDA: Origin Destination Surveys
 O&D and Mode Split Surveys OD&交通方式调查
 Different Methodologies.
 License Plate Study (video, audio, or handwritten)
 Check-Point / Intercept Interviews.
 Ask questions.在街上采访
 This can be at say, toll booth, or just on the street too.
 Video and Analysis. Observe. 视频和分析。守
 “Big Data” and BlueTooth “大数据”和蓝牙
 Phone Calls. Ask direct questions.拨打电话。提问
 Interpretation and Summary of Data
 Currently, Big Data has highest level of assumptions built into it,
because of:
 Limited data set comprised of cell phone users, or those that use blue tooth.
 Latency of up to a minute.
License Plate Survey, VIDEO
can “zoom in” later for clarity, engineer
just shoots video… easy
?
?
?
?
???
Can’t see all plates… this is a problem.
20 video Cameras Survey
…filmed during PEAK HOUR. Got ALL plates.
License Plate Survey, AUDIO
Requires visual clarity, can be difficult
for engineer to see…
?
?
?
???
Again, Can’t see many plates… Problem.
??
One Solution: Expensive.
 When you want to get ALL cars, you need to put
a video camera on ALL lanes, and get EVERY car.
 You must use high speed camera shutter.
 You need an overhead bridge, or some other way
to get a clear shot
 You need a camera operator for EACH lane.
 Even with only two locations, you need 6-8
camera operators and 8 cameras total.
 One such study for one day cost $45k in USA,
and had only 6 locations (25 camera operators).
Another Solution: Sample.
 A sample is different than a strict license plate survey.
 A sample gives you percentages of what direction people
are going, but it is with a RANDOM selection.
 Random selection can be done with watching video in
office and choosing a car.
 Random selection can NOT be done by picking a license
plate, because it biases depending on the lane visible to the
engineer. Far lanes are difficult to see, and eliminating
those from the sample creates a bias that is unknown.
 Also, because the license plate method requires matching
of plate data at each end, if there are too many missing
plates in the sampling, this makes it less reliable. Hard to
quantify if there are many entrance and exit points being
surveyed.
Origin Destination
let’s watch the Green Truck
人兴路 RENXING RD
VIDEO Origin & Destination SURVEY
A REALWORLD EXAMPLE:
Traffic Impact Analysis 交通影响分析
Project Description: Study Area
黄山大道
NEW ROAD? What Will Happen?
WB Traffic from JinKai
Traffic Will SHIFT with New Road
We had to define this
Traffic with an O&D Survey.
1. The traffic on the bridge was counted: 1630
2. Percentage of bridge to Freeway counted: 55%
Video Survey Corridors
Video Camera Tripod Locations
Selected Study Intersections
TOOK NEW TRAFFIC COUNTS
 Used iPhoneTraffic Count
Tool
 Used ManualTraffic Count
Sheet Method
Summary of Origin/Destination
 Traffic Count Data was used with O&D
Data
 Sample Size (100 vehicles or more) of
randomly selected vehicles that turned
left to go south at
Interchange/Intersection 8.
 In this Survey:
 55% of samples went south past机场高速
 45% turned right onto黄山大道
 Total LeftTurnVol = 1630 vph
 Result: 900 vehicles (55%) go south
 730 vehicles go west on黄山大道
Video Inspection Methodology
 In the video clips that follow, the
methodology used to sample traffic is set
forth.
 Video clips are “scrubbed” to watch a vehicle
very quickly, to see where it came from, and
to see where it went.
Origin Destination Samples
“Video Scrubbing Method”
Worksheet: Random Sampling of Vehicles
WB Traffic from JinKai
 PM Peak Hour is 2273 vehicles WB/SB on
金开大道. 1630 turns to go south on
人和大道. 643 goes west on 金开大道.
 55% of 1630 goes south on 人和大道
based on the O&DVideo Survey results.
 This is the control volume: 900
 45% of 1630 goes west on 黄山大道
Reassign Traffic from JinKai
NB RenHe to EB JinKai SHIFT?
Existing NB RenHe Traffic
Same Method Used
for NB Traffic: sampling
 Using Sampling methodology
lets the engineer multiple
percentage (%) values by the
TRAFFIC COUNT data to get
the amount of traffic that
would SHIFT.
 You need a statistically valid
sample size. At least 100
random samples.
Even the NB LEFT TURN traffic
is expected to SHIFT
Other Emerging Technologies
that can be used for O&D…
 BLUETOOTH
 How it works
 Advantages:
 It is time, date and location data, but needs interpretation
(bus passengers, bikes, peds, etc., mixed with car’s data)
 Can helps in determining paths and times
 Can be useful in timing studies, speed studies,O&D
 Disadvantages:
 It is just small sampling, not a traffic count
 Traffic counts still need to be performed.
 It is ONLY counting the cell phones that have blue tooth
turned ON
 Demographics of local Bluetooth users is not known
 Interpretation of data to determine car status, ped status,
bike, or even bus status…
Bluetooth Sampling Devices
can detect BTA phones, BTA computers, some cars
Bluetooth Sampling Technology
grabs a unique BT ID Address, and matches them
Bluetooth Sampling Technology
matches IDs at different measuring stations, and
computes travel times & speeds.
Bluetooth O&D Result Matches
Can also detect the path traveled (assumed),
depending on # of stations used. It’s still JUST
sampling… because not everyone uses bluetooth.
This kind of information is extremely helpful in setting up a travel
demand model, and calibrating it to real world conditions.
It is the trip distribution variable(s).
TRIP GENERATION SURVEYS
An Overview of
Trip Generation in USA
two thick manuals, 1500 pages,
users guidebook too.
Trip Generation in USA
 ITE Manual has 162 landuses
 China Manual has 36 landuses
 China manual can use more
categories and data.
Trip Generation Data in China is 5 times less than
what USA is using. More data is needed.
ITE Manual China Manual
162 landuses 36 landuses
ITE ITE ITE CHINA
ITE ITE
ITE says:
ITE Updates Trip Gen Manual
 Banking Industry, for example…
 ITE examined trip data for banks before and after
technology opened up ATMs and now Internet
Banking.
 Vehicle and Ped trips to the bank have gone down
significantly.
 Because of technology advances
 OlderTrip data for Banks (pre 2000) have been
removed to prevent skewing of data.
 Older data is discarded as our world changes.
 China is changing fast in many ways with vehicle
ownership UP, and large expansion of METRO lines.
ITE FORM only itemizes forTRUCKS. No PEDS. Mindset. Demographic
In China we also need to Count PEDESTRIANS and other MODES
Why would we need a
Trip Generation Survey?
 Lacking Data
 The ChinaTG Manual does not have rates for many land use
categories (such asVILLA, which is not in there)
 We,TYLIN, conducted aVILLA count because LongHu Project
hadVillas as a part of project.
 Villas have much higher vehicle trip counts than apts.
 Data Not Local. From another area of China
 Is it applicable?You will not know until you conduct aTG
Survey to find out if there are similarities.
 Are Demographics different?
 The Client has a New Land UseType, or a Hybrid
 Maybe the ChinaTG Manual does NOT have a hybrid but an
existing hybrid land use exists, and can be surveyed…
 If so, DO IT !
How to Scope a
Trip Generation Survey
 Define the Area to be surveyed
 The area should be a single land use type, such as
residential, or commercial, or institutional, etc.
 Identify all exit and entrance points.
 Trip Purpose:
 Find out the MODE SPLIT (car, ped, bike, etc)
 Find out theTrip Purpose (traffic model purposes)
 Training of Staff.
 Diagrams to help define specific duty.
 Photos to familiarize staff with area
 Execution coordinated by clock, overseen by supervisor
Example Project:
 礼嘉A区龙湖地块(A61-68)交通研究
 We had BeiJingTrip Rates already… but
 What should we use forTrip Generation?
 Since there is a residential apartment complex in the
City of ChongQing that was also developed by Long Hu
company (龙湖西苑 Long Hu Xi Juan), this is ideal and
similar. There are alsoVILLAS nearby across the street
 Find out pedestrian traffic
 Find out auto traffic
 This can be used forTraffic Study Project
 Also the 美地 MeiDiVillas development was surveyed
to get a sample of northern CQVilla trip generation.
PM Peak Person Trips / 100户数
(小户型)
Source:
ITE does not have PED data
PM Peak Person Trips / 100户数
(中户型)
Source:
ITE does not have PED data
PM Peak Person Trips / 100户数
(大户型)
Source:
ITE does not have PED data
PM Peak AUTO Trips / 100户数
(小户型)
Source:
PM Peak AUTO Trips / 100户数
(中户型)
Source:
PM Peak AUTO Trips / 100户数
(大户型)
Source:
Example: RESIDENTIAL Landuse
 Area in red is a
“closed” land use
area, gated, and filled
with 1683 residential
apartment units.
 There are several
locations to count
outside the fence
 Car entry/exit
 Pedestrian entry/exit
1683
DU
Example: RESIDENTIAL Landuse
1
3
5
4
4
2
 LongHuXiJuan
 Car Entry/Exit
 Ped Entry/Exit
 Also conduct an
O&D interview
 These areas all need
to be counted.
 6 locations.
 6 people.
Variables We Need to Know:
1
3
5 2
 how many cars total
(count at night,
garage, and street at
say, 2 am)
 population
 parking spaces
 shopping internal?
 other amenities
internal?
 Knowing these
variables lets us refine
ourTG information,
and even determine
parking needs, etc.
1683
DU
Count Location Photos
1
3
5
4
4
2
COUNT PEDESTRIANS
AND CARSTHAT GO IN
AND OUT OFTHESE
AUTO AND PED GATES,
FOR 3 HOURS
Count Location Photos
3
25
4
4
1
COUNT PEDESTRIANS
THAT GO UP AND
DOWNTHESE STAIRS
INTO LONGHUXIJUAN,
FOR 3 HOURS
Count Location Photos
3
1
5
4
4
2
COUNT PEDESTRIANS
AND CARSTHAT GO IN
AND OUT OFTHESE
AUTO AND PED GATES,
FOR 3 HOURS
Count Location Photos
3
5
4
4
2
1
COUNT PEDESTRIANS
THAT GOTO AND FROM
THIS SIDEWALK, FROM
LONGHUXIJUAN
Count Location Photos
3
1
5
4
4
2
COUNT PEDESTRIANS
THAT GO IN AND OUT
OFTHESE GATES FOR
3 HOURS (and how many
just go “through”)
Taking the Count: Location 1
3
1
5
4
4
2
COUNT PEDESTRIANS
AND CARS FORALL 6
MOVEMENTSON
DIAGRAM, FOR 3 HRS.
Peak hour:
5:15 pm-
6:15 pm
What We Get From Count:
 We get the total PEDESTRIANTraffic
that travels outside of the control area,
and “impacts” the street network and
sidewalks
 It is unknown how much of this traffic visits the
local shops and restaurants, but likely is a high
source.
 We get the total AUTOTraffic that
travels outside of the control area, and
“impacts” the street network.
 What we don’t know is how many
LongHuXiJuan residents who own cars, also
park their cars on the local street instead of the
internal garage. The streets are filled with
parked cars at night. It is likely that our “true”
AUTO trip generation rate is LOW in light of this.
Peds & Cars
COUNT RESULTS: Location 2
Peak
hour is
5:15-
6:15 pm
COUNT RESULTS: Location 4
Peak
hour is
5:00-
6:00 pm
COUNT RESULTS: Location 5
Peak
hour is
5:15-
6:15 pm
Comparing New with the OldPEDAUTO
MeiDi was higher, so
we used it
MeiDi was lower, so
we did not use it
LongHuXiJuan was
similar, so we used it
LongHuXiJuan was
similar, so we used it
Pedestrian Trip Generation Rates
Automobile Trip Generation Rates

More Related Content

PPTX
Trip generation
DOCX
Road Junctions and Intersections
PPTX
Types of parking studies
PPTX
TRAFFIC VOLUME STUDIES
PPT
Traffic Volume Study
PPTX
Integrated Public Transport System - Bangalore
PDF
Traffic & Transportation surveys
PPTX
report on parking. presentation
Trip generation
Road Junctions and Intersections
Types of parking studies
TRAFFIC VOLUME STUDIES
Traffic Volume Study
Integrated Public Transport System - Bangalore
Traffic & Transportation surveys
report on parking. presentation

What's hot (20)

PPTX
Transportation planning
PDF
L20 Weaving Merging and Diverging Movements
PDF
Geometric Design of Roads and Highways
PPTX
Origin – Destination survey
PDF
Lecture 03 Intersection Control (Traffic Engineering هندسة المرور & Dr. Usama...
PPT
4-Vertical and Horizontal Alignment Coordination(by Malyar Talash: Highway De...
PPT
Ahmedabad BRTS Overview
PPTX
Traffic Engineering And Drainage
PPTX
TRAFFIC SURVEYS- TRANSPORTATION ENGINEERING
PPTX
Transit Oriented Development.
PPTX
Traffic volume survey
PPTX
Comprehensive mobility plan jaipur
PDF
Chapter 3 characteristics of highway components
PPTX
TRANSIT ORIENTED DEVELOPMENT
PPT
Trip generation
PDF
Evolution of planning legislation in india
PPTX
Origin & destination survey
PPT
Chapter 1-road-cross-section-elements
PPT
Off street parking
PPTX
Land-use transport planning
Transportation planning
L20 Weaving Merging and Diverging Movements
Geometric Design of Roads and Highways
Origin – Destination survey
Lecture 03 Intersection Control (Traffic Engineering هندسة المرور & Dr. Usama...
4-Vertical and Horizontal Alignment Coordination(by Malyar Talash: Highway De...
Ahmedabad BRTS Overview
Traffic Engineering And Drainage
TRAFFIC SURVEYS- TRANSPORTATION ENGINEERING
Transit Oriented Development.
Traffic volume survey
Comprehensive mobility plan jaipur
Chapter 3 characteristics of highway components
TRANSIT ORIENTED DEVELOPMENT
Trip generation
Evolution of planning legislation in india
Origin & destination survey
Chapter 1-road-cross-section-elements
Off street parking
Land-use transport planning
Ad

Viewers also liked (20)

PDF
3-Trip Generation-Distribution ( Transportation and Traffic Engineering Dr. S...
PPTX
TRANSPORTATION PLANNING
PPT
O & d survey
DOCX
origin and destination survey research papeer
PPTX
O and d study
PPTX
Modelling car trip generation in the developing world the tale of two cities
PPTX
Traffic & transportation – ii
PDF
Lec 14 Demand Forecasting 2008 (Transportation Engineering Dr.Lina Shbeeb)
PPT
Urban transportation planning
PDF
Determining trip generation of commercial land use of kaptai
PPT
04 transport modelling
PPT
Survey on-pollution-1
PPTX
Transportation Survey
PPTX
land use by lowry model as an example
PPT
Presentation on Traffic Volume Survey
PDF
Spot speed studies and speed delay time survey
PPTX
Urban transportation system - methods of route assignment
PDF
Traffic and Transportation problems
PPTX
Traffic Volume Studies
PPTX
4.10.2 write models with calc reg
3-Trip Generation-Distribution ( Transportation and Traffic Engineering Dr. S...
TRANSPORTATION PLANNING
O & d survey
origin and destination survey research papeer
O and d study
Modelling car trip generation in the developing world the tale of two cities
Traffic & transportation – ii
Lec 14 Demand Forecasting 2008 (Transportation Engineering Dr.Lina Shbeeb)
Urban transportation planning
Determining trip generation of commercial land use of kaptai
04 transport modelling
Survey on-pollution-1
Transportation Survey
land use by lowry model as an example
Presentation on Traffic Volume Survey
Spot speed studies and speed delay time survey
Urban transportation system - methods of route assignment
Traffic and Transportation problems
Traffic Volume Studies
4.10.2 write models with calc reg
Ad

Similar to TRIP GENERATION Survey and ORIGIN / DESTINATION STUDY (20)

DOCX
GeoTeam_GT-150-13_Data Processing Report_Rev01
PDF
Contoh Desain Slide Presentasi Ilmiah Kreatif dan Menarik #1
PDF
Mounir_Sardouk_Final_Report_touch
PDF
IRJET- Robust and Fast Detection of Moving Vechiles in Aerial Videos usin...
PPT
VIRDI COMPANY PROFILE & PRODUCTS OVERVIEW
PPTX
Using Crowdsourcing, Automated Methods and Google Street View to Collect Side...
PPTX
London web performance WPO Lessons from the field June 2013
PDF
Advanced FPGA Designs Tools
PPS
Fujitsu APD Introduction
PDF
Contoh Desain Slide Presentasi Ilmiah Kreatif dan Menarik #3
PDF
COMPARATIVE STUDY ON AUTOMATED NUMBER PLATE EXTRACTION USING OPEN CV AND MATLAB
PPTX
Multi-Resolution Modeling by using Vision Traffic Suite
PDF
IRJET- Video Based Traffic Sign Detection by Scale Based Frame Fusion Technique
PDF
License Plate Recognition
PDF
OPTIMIZING DRILLING & WELLBORE POSITIONING THROUGH AUTOMATION AND DIGITALIZATION
PDF
Build and Host Real-world Machine Learning Services from Scratch @ pycontw2019
DOCX
Project report for dgm 1
PDF
Disaster porn and the value of a generalist
GeoTeam_GT-150-13_Data Processing Report_Rev01
Contoh Desain Slide Presentasi Ilmiah Kreatif dan Menarik #1
Mounir_Sardouk_Final_Report_touch
IRJET- Robust and Fast Detection of Moving Vechiles in Aerial Videos usin...
VIRDI COMPANY PROFILE & PRODUCTS OVERVIEW
Using Crowdsourcing, Automated Methods and Google Street View to Collect Side...
London web performance WPO Lessons from the field June 2013
Advanced FPGA Designs Tools
Fujitsu APD Introduction
Contoh Desain Slide Presentasi Ilmiah Kreatif dan Menarik #3
COMPARATIVE STUDY ON AUTOMATED NUMBER PLATE EXTRACTION USING OPEN CV AND MATLAB
Multi-Resolution Modeling by using Vision Traffic Suite
IRJET- Video Based Traffic Sign Detection by Scale Based Frame Fusion Technique
License Plate Recognition
OPTIMIZING DRILLING & WELLBORE POSITIONING THROUGH AUTOMATION AND DIGITALIZATION
Build and Host Real-world Machine Learning Services from Scratch @ pycontw2019
Project report for dgm 1
Disaster porn and the value of a generalist

Recently uploaded (20)

PPTX
Sustainable Sites - Green Building Construction
PPTX
Construction Project Organization Group 2.pptx
PDF
SM_6th-Sem__Cse_Internet-of-Things.pdf IOT
PDF
Operating System & Kernel Study Guide-1 - converted.pdf
PDF
Arduino robotics embedded978-1-4302-3184-4.pdf
PDF
The CXO Playbook 2025 – Future-Ready Strategies for C-Suite Leaders Cerebrai...
PPTX
Geodesy 1.pptx...............................................
PDF
Mitigating Risks through Effective Management for Enhancing Organizational Pe...
PPTX
OOP with Java - Java Introduction (Basics)
PPTX
Engineering Ethics, Safety and Environment [Autosaved] (1).pptx
PDF
Digital Logic Computer Design lecture notes
PPTX
Welding lecture in detail for understanding
PPTX
M Tech Sem 1 Civil Engineering Environmental Sciences.pptx
PDF
PPT on Performance Review to get promotions
PDF
Mohammad Mahdi Farshadian CV - Prospective PhD Student 2026
PPT
Mechanical Engineering MATERIALS Selection
PPTX
Lesson 3_Tessellation.pptx finite Mathematics
PDF
keyrequirementskkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkk
PPTX
Lecture Notes Electrical Wiring System Components
PPTX
IOT PPTs Week 10 Lecture Material.pptx of NPTEL Smart Cities contd
Sustainable Sites - Green Building Construction
Construction Project Organization Group 2.pptx
SM_6th-Sem__Cse_Internet-of-Things.pdf IOT
Operating System & Kernel Study Guide-1 - converted.pdf
Arduino robotics embedded978-1-4302-3184-4.pdf
The CXO Playbook 2025 – Future-Ready Strategies for C-Suite Leaders Cerebrai...
Geodesy 1.pptx...............................................
Mitigating Risks through Effective Management for Enhancing Organizational Pe...
OOP with Java - Java Introduction (Basics)
Engineering Ethics, Safety and Environment [Autosaved] (1).pptx
Digital Logic Computer Design lecture notes
Welding lecture in detail for understanding
M Tech Sem 1 Civil Engineering Environmental Sciences.pptx
PPT on Performance Review to get promotions
Mohammad Mahdi Farshadian CV - Prospective PhD Student 2026
Mechanical Engineering MATERIALS Selection
Lesson 3_Tessellation.pptx finite Mathematics
keyrequirementskkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkk
Lecture Notes Electrical Wiring System Components
IOT PPTs Week 10 Lecture Material.pptx of NPTEL Smart Cities contd

TRIP GENERATION Survey and ORIGIN / DESTINATION STUDY

  • 1. Origin Destination Surveys OD调查 …and Trip Generation Surveys… 交通出行率调查
  • 2. Project Description 项 目概况 Project Setting 项目区域现状 Travel Demand Analysis 交通需求分析 Transportation Impact Analysis 交通影响分析 Transportation Mitigation Measures 交通改善措施 Appendix A, Appendix B, etc. etc. 附录A, 附录B, 等等
  • 3. But first…但首先, Traffic Counts, REVISITED… 高峰时段交叉口交通流量调查,回访
  • 4. Scoping Site in Advance for Intersection Traffic Counts 计算前的准备工作  Visit the project site  take pictures of traffic controls and lanes 拍摄交通控制 和路面情况  observe intersection conditions  Minimize Complications in taking traffic counts  Engineer should make determination of staff needs, in advance.  Timing of intervals, Recording  Complications means LOST DATA and a bad count on the day of survey.  Incorrect Counts  Missing Information.Wrong North Arrow, etc
  • 5. Traffic Count Sheet: Hard to Interpret 交通流量调查问卷: 用中文很难解释 Problem: Intersection mis- labeled because road closure was at LongShouLu
  • 6. Traffic Count Sheet: Incorrect Method 交通流量调查问卷: 错误的方法 Counted first... 5:30-5:45 pm Counted second... 5:45-6:00 pm Counted third... 6:00-6:15 pm
  • 7. Traffic Count Sheet: Hard to Interpret 交通流量调查问卷:用中文 很难解释
  • 8. Traffic Count Sheet: Only 15 Minute Count 交通流量调查问卷: 只有15分钟的流量调查
  • 9. 15 Minute Count is Not Enough 15分钟流量调查不够 Left Out Right Out Left In Right In 3:30 12 8 16 12 3:45 20 16 24 28 4:00 20 4 8 24 4:15 32 28 12 24 4:30 32 8 20 48 4:45 16 16 24 52 5:00 28 12 32 64 5:15 32 20 32 88 5:30 40 24 28 92 5:45 16 12 72 104 6:00 40 12 60 116 6:15 32 8 20 60 What you get when you multiply a 15 minute count by 4: DIFFERENT RESULTS! 不同的结果 32 17 48 100 32 17 48 100 32 17 48 100 32 17 48 100 32 17 48 100 32 17 48 100 32 17 48 100 32 17 48 100 32 17 48 100 32 17 48 100 32 17 48 100 Actual: Actual: Actual: Actual: Actual: Actual: Actual: Actual: Actual: Actual: Actual:
  • 10. 15 Minute Count is Not Enough 15分钟流量调查不够  Long Hu Xi Juan 3 hr. Count. 龙湖西苑3小时流量调查 A real count.  What happens when you multiply a 15 minute count by 4 to “short-cut” the count process. 4当你计算15分 钟的流量,会发生什么?You get vastly different hourly totals!你会得到不同的结果 Likely Incorrect. 很有可能是不正确的 0 20 40 60 80 100 120 140 3:30 3:45 4:00 4:15 4:30 4:45 5:00 5:15 5:30 5:45 6:00 6:15 Actual Left Out Right Out Left In Right In Actual Peak HourTotal实际高峰时段总 Actual Peak HourTotal 实际高峰时段总 Actual Peak HourTotal实际高峰时段总 PeakHourVolume 左出口 右出口 左入口 右入口
  • 11. We need to Count for:  TWO HOURS minimum  We find the peak HOUR from these two hours of data.  It cannot be guessed…  …it is incorrect to multiply a 15 minute count by four(4) to get an hourly total. 15min x 4 peak hour
  • 12. 15 Minute Count is Not Enough  This example shows that you get four very different answers  The first 15 minutes x 4 is 15% of actual.  The second 15 minutes x 4 is 25% of actual.  The third 15 minutes x 4 is 50% of actual, etc. etc.  The fourth set was close, but that was LUCKY 0 20 40 60 80 100 120 140 3:30 3:45 4:00 4:15 4:30 4:45 5:00 5:15 5:30 5:45 6:00 6:15 Actual Left Out Right Out Left In Right In Actual Peak HourTotal Actual Peak HourTotal Actual Peak HourTotal PeakHourVolume
  • 13. Photo Documentation is very Important and Helpful 龙寿路 /人兴路 LongShou / RenXing
  • 14. Photo Location 1: 1 龙寿路 /人兴路 LongShou / RenXing 1
  • 15. Photo Location 1: 2 龙寿路 /人兴路 LongShou / RenXing 1
  • 16. Photo Location 1: 3 龙寿路 /人兴路 LongShou / RenXing 1
  • 17. Photo Location 2: 1 龙寿路 /人兴路 LongShou / RenXing 2
  • 18. Photo Location 2: 2 龙寿路 /人兴路 LongShou / RenXing 2
  • 19. Photo Location 3: 1 龙寿路 /人兴路 LongShou / RenXing 3
  • 20. Photo Location 3: 2 龙寿路 /人兴路 LongShou / RenXing 3
  • 21. Photo Location 4: 1 龙寿路 /人兴路 LongShou / RenXing 4
  • 23. TDA: Origin Destination Surveys  O&D and Mode Split Surveys OD&交通方式调查  Different Methodologies.  License Plate Study (video, audio, or handwritten)  Check-Point / Intercept Interviews.  Ask questions.在街上采访  This can be at say, toll booth, or just on the street too.  Video and Analysis. Observe. 视频和分析。守  “Big Data” and BlueTooth “大数据”和蓝牙  Phone Calls. Ask direct questions.拨打电话。提问  Interpretation and Summary of Data  Currently, Big Data has highest level of assumptions built into it, because of:  Limited data set comprised of cell phone users, or those that use blue tooth.  Latency of up to a minute.
  • 24. License Plate Survey, VIDEO can “zoom in” later for clarity, engineer just shoots video… easy ? ? ? ? ??? Can’t see all plates… this is a problem.
  • 25. 20 video Cameras Survey …filmed during PEAK HOUR. Got ALL plates.
  • 26. License Plate Survey, AUDIO Requires visual clarity, can be difficult for engineer to see… ? ? ? ??? Again, Can’t see many plates… Problem. ??
  • 27. One Solution: Expensive.  When you want to get ALL cars, you need to put a video camera on ALL lanes, and get EVERY car.  You must use high speed camera shutter.  You need an overhead bridge, or some other way to get a clear shot  You need a camera operator for EACH lane.  Even with only two locations, you need 6-8 camera operators and 8 cameras total.  One such study for one day cost $45k in USA, and had only 6 locations (25 camera operators).
  • 28. Another Solution: Sample.  A sample is different than a strict license plate survey.  A sample gives you percentages of what direction people are going, but it is with a RANDOM selection.  Random selection can be done with watching video in office and choosing a car.  Random selection can NOT be done by picking a license plate, because it biases depending on the lane visible to the engineer. Far lanes are difficult to see, and eliminating those from the sample creates a bias that is unknown.  Also, because the license plate method requires matching of plate data at each end, if there are too many missing plates in the sampling, this makes it less reliable. Hard to quantify if there are many entrance and exit points being surveyed.
  • 30. 人兴路 RENXING RD VIDEO Origin & Destination SURVEY A REALWORLD EXAMPLE: Traffic Impact Analysis 交通影响分析
  • 31. Project Description: Study Area 黄山大道
  • 32. NEW ROAD? What Will Happen?
  • 33. WB Traffic from JinKai
  • 34. Traffic Will SHIFT with New Road
  • 35. We had to define this Traffic with an O&D Survey. 1. The traffic on the bridge was counted: 1630 2. Percentage of bridge to Freeway counted: 55%
  • 37. Video Camera Tripod Locations
  • 39. TOOK NEW TRAFFIC COUNTS  Used iPhoneTraffic Count Tool  Used ManualTraffic Count Sheet Method
  • 40. Summary of Origin/Destination  Traffic Count Data was used with O&D Data  Sample Size (100 vehicles or more) of randomly selected vehicles that turned left to go south at Interchange/Intersection 8.  In this Survey:  55% of samples went south past机场高速  45% turned right onto黄山大道  Total LeftTurnVol = 1630 vph  Result: 900 vehicles (55%) go south  730 vehicles go west on黄山大道
  • 41. Video Inspection Methodology  In the video clips that follow, the methodology used to sample traffic is set forth.  Video clips are “scrubbed” to watch a vehicle very quickly, to see where it came from, and to see where it went.
  • 42. Origin Destination Samples “Video Scrubbing Method”
  • 44. WB Traffic from JinKai  PM Peak Hour is 2273 vehicles WB/SB on 金开大道. 1630 turns to go south on 人和大道. 643 goes west on 金开大道.  55% of 1630 goes south on 人和大道 based on the O&DVideo Survey results.  This is the control volume: 900  45% of 1630 goes west on 黄山大道
  • 46. NB RenHe to EB JinKai SHIFT?
  • 47. Existing NB RenHe Traffic
  • 48. Same Method Used for NB Traffic: sampling  Using Sampling methodology lets the engineer multiple percentage (%) values by the TRAFFIC COUNT data to get the amount of traffic that would SHIFT.  You need a statistically valid sample size. At least 100 random samples.
  • 49. Even the NB LEFT TURN traffic is expected to SHIFT
  • 50. Other Emerging Technologies that can be used for O&D…  BLUETOOTH  How it works  Advantages:  It is time, date and location data, but needs interpretation (bus passengers, bikes, peds, etc., mixed with car’s data)  Can helps in determining paths and times  Can be useful in timing studies, speed studies,O&D  Disadvantages:  It is just small sampling, not a traffic count  Traffic counts still need to be performed.  It is ONLY counting the cell phones that have blue tooth turned ON  Demographics of local Bluetooth users is not known  Interpretation of data to determine car status, ped status, bike, or even bus status…
  • 51. Bluetooth Sampling Devices can detect BTA phones, BTA computers, some cars
  • 52. Bluetooth Sampling Technology grabs a unique BT ID Address, and matches them
  • 53. Bluetooth Sampling Technology matches IDs at different measuring stations, and computes travel times & speeds.
  • 54. Bluetooth O&D Result Matches Can also detect the path traveled (assumed), depending on # of stations used. It’s still JUST sampling… because not everyone uses bluetooth. This kind of information is extremely helpful in setting up a travel demand model, and calibrating it to real world conditions. It is the trip distribution variable(s).
  • 56. Trip Generation in USA two thick manuals, 1500 pages, users guidebook too.
  • 57. Trip Generation in USA  ITE Manual has 162 landuses  China Manual has 36 landuses  China manual can use more categories and data.
  • 58. Trip Generation Data in China is 5 times less than what USA is using. More data is needed. ITE Manual China Manual 162 landuses 36 landuses ITE ITE ITE CHINA ITE ITE ITE says:
  • 59. ITE Updates Trip Gen Manual  Banking Industry, for example…  ITE examined trip data for banks before and after technology opened up ATMs and now Internet Banking.  Vehicle and Ped trips to the bank have gone down significantly.  Because of technology advances  OlderTrip data for Banks (pre 2000) have been removed to prevent skewing of data.  Older data is discarded as our world changes.  China is changing fast in many ways with vehicle ownership UP, and large expansion of METRO lines.
  • 60. ITE FORM only itemizes forTRUCKS. No PEDS. Mindset. Demographic In China we also need to Count PEDESTRIANS and other MODES
  • 61. Why would we need a Trip Generation Survey?  Lacking Data  The ChinaTG Manual does not have rates for many land use categories (such asVILLA, which is not in there)  We,TYLIN, conducted aVILLA count because LongHu Project hadVillas as a part of project.  Villas have much higher vehicle trip counts than apts.  Data Not Local. From another area of China  Is it applicable?You will not know until you conduct aTG Survey to find out if there are similarities.  Are Demographics different?  The Client has a New Land UseType, or a Hybrid  Maybe the ChinaTG Manual does NOT have a hybrid but an existing hybrid land use exists, and can be surveyed…  If so, DO IT !
  • 62. How to Scope a Trip Generation Survey  Define the Area to be surveyed  The area should be a single land use type, such as residential, or commercial, or institutional, etc.  Identify all exit and entrance points.  Trip Purpose:  Find out the MODE SPLIT (car, ped, bike, etc)  Find out theTrip Purpose (traffic model purposes)  Training of Staff.  Diagrams to help define specific duty.  Photos to familiarize staff with area  Execution coordinated by clock, overseen by supervisor
  • 63. Example Project:  礼嘉A区龙湖地块(A61-68)交通研究  We had BeiJingTrip Rates already… but  What should we use forTrip Generation?  Since there is a residential apartment complex in the City of ChongQing that was also developed by Long Hu company (龙湖西苑 Long Hu Xi Juan), this is ideal and similar. There are alsoVILLAS nearby across the street  Find out pedestrian traffic  Find out auto traffic  This can be used forTraffic Study Project  Also the 美地 MeiDiVillas development was surveyed to get a sample of northern CQVilla trip generation.
  • 64. PM Peak Person Trips / 100户数 (小户型) Source: ITE does not have PED data
  • 65. PM Peak Person Trips / 100户数 (中户型) Source: ITE does not have PED data
  • 66. PM Peak Person Trips / 100户数 (大户型) Source: ITE does not have PED data
  • 67. PM Peak AUTO Trips / 100户数 (小户型) Source:
  • 68. PM Peak AUTO Trips / 100户数 (中户型) Source:
  • 69. PM Peak AUTO Trips / 100户数 (大户型) Source:
  • 70. Example: RESIDENTIAL Landuse  Area in red is a “closed” land use area, gated, and filled with 1683 residential apartment units.  There are several locations to count outside the fence  Car entry/exit  Pedestrian entry/exit 1683 DU
  • 71. Example: RESIDENTIAL Landuse 1 3 5 4 4 2  LongHuXiJuan  Car Entry/Exit  Ped Entry/Exit  Also conduct an O&D interview  These areas all need to be counted.  6 locations.  6 people.
  • 72. Variables We Need to Know: 1 3 5 2  how many cars total (count at night, garage, and street at say, 2 am)  population  parking spaces  shopping internal?  other amenities internal?  Knowing these variables lets us refine ourTG information, and even determine parking needs, etc. 1683 DU
  • 73. Count Location Photos 1 3 5 4 4 2 COUNT PEDESTRIANS AND CARSTHAT GO IN AND OUT OFTHESE AUTO AND PED GATES, FOR 3 HOURS
  • 74. Count Location Photos 3 25 4 4 1 COUNT PEDESTRIANS THAT GO UP AND DOWNTHESE STAIRS INTO LONGHUXIJUAN, FOR 3 HOURS
  • 75. Count Location Photos 3 1 5 4 4 2 COUNT PEDESTRIANS AND CARSTHAT GO IN AND OUT OFTHESE AUTO AND PED GATES, FOR 3 HOURS
  • 76. Count Location Photos 3 5 4 4 2 1 COUNT PEDESTRIANS THAT GOTO AND FROM THIS SIDEWALK, FROM LONGHUXIJUAN
  • 77. Count Location Photos 3 1 5 4 4 2 COUNT PEDESTRIANS THAT GO IN AND OUT OFTHESE GATES FOR 3 HOURS (and how many just go “through”)
  • 78. Taking the Count: Location 1 3 1 5 4 4 2 COUNT PEDESTRIANS AND CARS FORALL 6 MOVEMENTSON DIAGRAM, FOR 3 HRS. Peak hour: 5:15 pm- 6:15 pm
  • 79. What We Get From Count:  We get the total PEDESTRIANTraffic that travels outside of the control area, and “impacts” the street network and sidewalks  It is unknown how much of this traffic visits the local shops and restaurants, but likely is a high source.  We get the total AUTOTraffic that travels outside of the control area, and “impacts” the street network.  What we don’t know is how many LongHuXiJuan residents who own cars, also park their cars on the local street instead of the internal garage. The streets are filled with parked cars at night. It is likely that our “true” AUTO trip generation rate is LOW in light of this. Peds & Cars
  • 80. COUNT RESULTS: Location 2 Peak hour is 5:15- 6:15 pm
  • 81. COUNT RESULTS: Location 4 Peak hour is 5:00- 6:00 pm
  • 82. COUNT RESULTS: Location 5 Peak hour is 5:15- 6:15 pm
  • 83. Comparing New with the OldPEDAUTO MeiDi was higher, so we used it MeiDi was lower, so we did not use it LongHuXiJuan was similar, so we used it LongHuXiJuan was similar, so we used it

Editor's Notes

  • #2: Origin Destination Surveys OD调查 diào chá investigation; inquiry; to investigate; to survey; survey; (opinion) poll; CL:項|项[xiang4],個|个[ge4] HSK4
  • #3: Today’s presentation has to do with the third element of the Traffic Impact Analysis, Travel Demand Analysis. It has to do with collecting and developing the data inputs for the transportation impact analyses.
  • #4: Before we proceed with the presentation on surveys, I want to review what we talked about before. I want to discuss traffic count data collection procedures.
  • #5: It is very important for the engineer in charge to first scope out the traffic count site. It is important to take note of any complications that might come up during the count. Pictures should be taken. Once an engineer receives training or has experience on what to do, they can be independent. Unless a professional company collects the data, it is common that even engineers make mistakes in how this is to be done. It is important to be familiar with: Knowing how many people are needed. What to do if there are complications. How to correct bad data collection.
  • #6: This sheet of paper was mislabeled, because WanShou was not counted. This mistake required interpretation at the office by the P.M.
  • #7: This count sheet shows three different counts, taken at the same intersection. The first count was for 15 minutes for the north direction. The second 15 minute count was for the east direction. The third 15 minute count was for the south direction. The problem is that this should have been a two hour count for ALL directions, at the same time. There should have been eight (8) 15 minute count sheets. There should have been 8 sets of 15 minute counts. This set of 3 15 minute counts, for different directions, was not useable, and had to be discarded.
  • #8: This count sheet was hard to read, hard to interpret. There was again incorrect labeling of street names, that had to be corrected in the office later, because it was the wrong street names. This leads to confusion.
  • #9: This count sheet shows a lack of street names, and so it is difficult for P.M. to interpret this, to make sure the data is correct. All data and directions, etc., should be clearly labeled in the field, so that no one has to guess in the office.
  • #10: A 15 minute count seems like it might be good enough. But it is not good enough. Not when we use average hourly count volumes in our software. These 15 minute counts are very different. This slide shows that multiplying a 15 minute count by 4, for any of the twelve 15 minute intervals, will give a very different answer than adding up the four highest consecutive 15 minute counts. Some of the differences are very significant. In this example, the peak hour was calculated to be from 5:15 pm to 6:15 pm. It is shown in the yellow block. The totals of the peak hour are shown at the bottom of the sheet, also in yellow.
  • #11: This chart shows a few examples of the results we get by multiplying a 15 minute count interval by 4. All six examples are very different.
  • #46: 900 cars now go south past the freeway in pm peak hour on RenHe Blvd. This is the only path currently crossing the freeway. However, the train station and other traffic on the major street TaiShan Blvd could attract 50% of this southbound traffic (since there is a 2nd road connector, it will likely get half). Of the traffic that will go south on RenHe and to the south and west, it is assumed that 50% of THAT traffic will also use the new RenXing extension connector to avoid the major congestion on RenHe southbound between JinKai and the Freeway. So, 75% of this traffic (900 vph) is expected to use the new road. This is 675 cars turning left to go on these local streets. This is a conservatively high estimate, assuming that traffic will shift to the less congested pathways over time. Contra Flow was assumed also, in proportion to the SB and NB split at Interchange Junction 8, where the total SB traffic from JinKai is 1630 and the NBR onto JinKai is 1500, therefore this is a close 50/50 split. What is not known, is how many of the current 1500 NBR came from the freeway interchange. It IS known that 900 cars went SB from JinKai to the interchange, by the OandD survey. In this study, we are going to assume 50% of the 1500 came from RenHe Blvd south of the freeway crossings. So use 750 for NB volume that will shift. Therefore 750 cars were shifted from RenHe NB to RenXing NB along these two paths. The result was LOS B/C at all RenXing intersections for Year 2013 conditions traffic levels.
  • #50: The opening of intersection 7 to signalization will also create a demand for a northbound RenXing to westbound JinKai Blvd. This new movement will require left turn signalization phasing as well. In Alternative 2 the fully accessible Intersection #7 was examined. which would allow all turning movements at this intersection. The amount of traffic going north onto the bridge from RenHe is 112 cars in 4 minutes. and this is 1680 vph. Say 50% of this is going straight, and 50% is going west or left onto JinKai. So 840 cars go left onto JinKai, and 840 cars go north on RenHe at the bridge gore. ASSUME that half of these will shift, or 420 more cars going north onto RenXing and making a left at Intersection #7 to go west onto JinKai as an alternative to the congested RenHe route. LOS C or better is result at all four (4) RenXing intersections even after these changes.