Botswana_Guideline 9 - Traffic Data Collection and Analysis (2004)

5/21/2018 Botswana_Guideline 9 - Traffic Data Collection and Analysis (2004)

1/54

1Traffic Data Collection and Analysis

Roads Department

Traffic Data Collectionand Analysis


2/54

2 Traffic Data Collection and Analysis

Roads Department

Ministry o Works and Transportoads Departmentrivate Bag 0026aborone, Botswana

hone + 7 - 5Fax + 2 7 - 14 27

February 2004

ISBN 99912 - 0 - 417 - 2

eproduction of extracts from this Guideline may be made subject to due acknowledgement of the source.

t oug t s Gu e ne s be eve to be correct at t e t me of pr nt ng, Roa s Department oes not accept any con-

tractual, tortious or other form of liability for its contents or for any consequences arising from its use. Anyone using

t e nformat on conta ne n t e Gu e ne s ou app y t e r own s an u gement to t e part cu ar ssue un er

consi eration.


3/54


Roads Department

ROADS DEPARTMENT

nder the policy direction of the Ministry of Works and Transport, Roads Department is responsible for provid-

ng an a equate, safe, cost-effect ve an ef c ent roa nfrastructure w t n t e bor ers of Botswanaas we as for

acilitating cross-border road communications with neighbouring countries. Implied in these far-ranging responsi-

b t es s t e ob gat on to:

. ensure t at ex st ng roa s are a equate y ma nta ne n or er to prov e appropr ate eve of serv ce for

roa users;

. mprove ex st ng roa s to requ re stan ar s to enab e t em to carry preva ng eve s of traf c w t t e

required degree of safety;

3. provide new roads to the required geometric, pavement design and safety standards.

he Department has been vested with the strategic responsibility for overall management of the Public HighwayNetwor PHN of some 18, 300 m of roa s. T s confers aut or ty for sett ng of nat ona spec cat ons an

stan ar s an s eare respons b ty w t t e D str ct Counc s an Department of W fe an Nat ona Par s for

the co-ordinated planning of the PHN.

oads Department is also responsible for administering the relevant sections of the Public Roads Act, assist-

ng oca roa aut or t es on tec n ca matters an prov ng ass stance n t e nat ona effort to promote c t zen

contractors in the road construction industry by giving technical advice wherever possible. This task is facilitate

by t e pub cat on of a ser es of Tec n ca Gu e nes ea ng w t stan ar s, genera proce ures an best pract ce

on a variety of aspects of the planning, design, construction and maintenance of roads in Botswana that take full

account of oca con t ons.

Guideline No. 1 The Design, Construction and Maintenance of Otta Seals (1999).Addendum to Guideline No. 1 Seminar Proceedings (June, 2000).

Guideline No. 2 Pavement Testing, Analysis and Interpretation of Test Data (2000).

Addendum to Guideline No. 2 Seminar Proceedings (January, 2002).

Guideline No. 3 Methods and Procedures for Prospecting for Road Construction Materials (2000).

Addendum to Guideline No. 3 Seminar Proceedings (April, 2002).

Guideline No. 4 Axle Load Surveys (2000).

Addendum to Guideline No. 4 Seminar Proceedings (January, 2002).

u del ne No. 5 Plann ng and Env ronmental Impact Assessment o Road In rastructure 2001 .

Guideline No. 6 The Prevention and Repair of Salt Damage to Roads and Runways (2001).

Addendum to Guideline No. 6 Seminar Proceedings (April, 2002)

Guideline No. 7 Technical Auditing of Road Projects (2001).

Addendum to Guideline No. 7 Seminar Proceedings (June, 2003).

Guideline No. 8 The use of Silcrete and Other Marginal Materials for Road Surfacing (2002).

Addendum to Guideline No. 8 Seminar Proceedings (June, 2003).

Guideline No. 9 Traffic Data Collection and Analysis (2003).


4/54

Traffic Data Collection and Analysis

Roads Department

FOREWORD

espite the different core areas of road use to which these Guidelines pertain, the ultimate objective is to ensure

proper, a equate, safe econom ca an ef c ent management of t e nat ona roa networ . It s wort po nt ng outthat, without physical access to obs, health, markets, education and other amenities, the quality of life inevitably

suffers, growth stagnates and poverty reduction cannot be sustained. These Guidelines, therefore, will enable

a opt on of a common approac an met o o ogy for co ect on of traf c ata y t e respect ve aut or t es, w e

at the same time ensuring that data so collected fits the purpose for which it is intended to serve.

he approach as provided for in these Guidelines, will enhance proactive action by planners, implementing bod-

es an ot er bene c ar es n prov ng t e necessary a v se to t e var ous aut or t es on matters of stat st ca traf-

c data collection. This will further enable appropriate and sustainable decisions to be reached by the authorities

concerne on w et er or not a new roa s requ re for t e preva ng traf c con t on, ascerta n t e t m ng of t e

requ re mprovements suc as capac ty expans on, structura upgra ng or strengt en ng , etc. To t s en , t e

following obligations are key to the provision and maintenance of the Public Highway Network:

. Provision of the maximum road benefit to the greater number of people for the minimum expenditure or cost

to t e economy, by concentrat ng a construct on an ma ntenance efforts on roa s carry ng t e eav er traf c

and maintaining its economic length.

. Provision of improved roads to areas of administrative and district centers with regard to the resulting eco-

nom c or soc a eve opment of t e area of popu at on.

. Strategic responsibility of managing the Public Highway Network with respect to the setting of specifications

an stan ar s, an prov s on of gu e nes n t e con uct of surveys an stu es.

. Ensure prov s on of surface /b tumen roa networ connect ons, t at s a weat er roa s to towns/c t es an

places of economic activities.

While taking cognisance of the above, the attention is also drawn to the observation that the Southern African

eve opment Commun ty SADC reg on s face w t a tremen ous c a enge n roa eve opment an susta n-

able maintenance of the existing infrastructure. This scenario is resulting in increased transport operating costs,

oss in road investment, etc., which is coupled with declining funds allocations earmarked for road provision an

ma ntenance. It s n v ew of t ese occurrences n t e reg on t at Botswana s ou cons er t cruc a to appropr -

ately collect, analyse and sustainably maintain sufficient traffic data as key input parameters in decision making

for eve opment an management of t e Pub c H g way Networ .

aborone

May 2003

Andrew Nkaro

irector of Roads

oa s Department

Ministry of Works and Transport


5/54


Roads Department

ACKNOWLEDGEMENTS

his Guideline is one of a series that is being produced under the Institutional Co-orperation Agreement that exists

between the Roads Department and the Norwegian Public Roads Administration (NPRA). This Agreement fallsun er a NORAD Tec n ca Ass stance Programme to Roa s Department, w c s co-fun e by t e K ng om o

Norway and the Government of the Republic of Botswana.

he production of this Guideline has been a joint effort between the Roads Department and the Norwegian Public

oa s A m n strat on NPRA . Messrs. Per Engeset of NPRA an Obus tswe B. Ke t e e of Botswana Roa s

epartment jointly compiled this Guideline.

a ue support, suggest ons an ar wor for t e pro uct on of t e Gu e ne t roug comments on var ous

rafts has come from various people, in particular; Messrs. Bjrn K re Steinset and Charles Overby, both from

Norweg an Pub c Roa s A m n strat on, an Dr. Bernar Ob a, Roug ton Internat ona , UK.

e fo ow ng n v ua s are a so cre te for prov ng t e r nputs/comments an ass stance at var ous stagesof compiling this Guideline:

r. B. K. Sharma Roads Department

s. Lorato Ts aa ane Roa s Department

s. Kgakgamatso Matu Roads Department

r. Diganta Sarma Roads Department

r. F. K. Nor u Gaborone C ty Counc

otograp s were prov e by:

r. Per Engeset - Norweg an Pub c Roa s A m n strat on

r. Charles Overby - Norwegian Public Roads Administration

r. O us tswe. B. Ke t e e Roa s Department

s. Kgakgamatso Matu Roads Department


6/54


Roads Department

TABLE OF CONTENTS

1 INTRODUCTION ............................................................................................................................................9

. ac groun.2 Purpose an Scope of t e Gu e nes.......................................................................................................

.3 Structure of the Guideline 1

2 TRAFFIC DATA COLLECTION ................................................................................................................12

.1 Role and Function

.2 Types of Traf c Counts ............................................................................................................................

2.2.1 Manual ounts

2.2.2 Automat c Counts .........................................................................................................................

.3 General Specifications..............................................................................................................................

3 TRAFFIC DATA COLLECTION, PRACTISES IN BOTSWANA ..........................................................17

.1 enera

3.2 Selection of Counting Sites ......................................................................................................................

3.3 Vehicle Classification and Configuration

COUNTING CHARACTERISTICS ............................................................................................................22

.1 Traf c F ow

.1.1 Frequency of Traffic Counts.........................................................................................................

.1.2 Data Acqu s t on

.1.3 Routine Counts

.1.4 A -Hoc or nsc e u e ounts...................................................................................................

.1.5 Origin Destination Surveys 6

A A A.................................................

.1 enera

.2 Staff Compos t on an Qua cat on ........................................................................................................

. qu pment equ rement

. . utomat c ra c ount ng qu pment .......................................................................................

6 TRAFFIC AND SITE SAFETY....................................................................................................................3

.1 Genera .....................................................................................................................................................

.2 Site Safety

.3 S te Mar ngs ...........................................................................................................................................

7 TRAFFIC COUNTING PROCEDURES ....................................................................................................31

7.1 Genera 1

7.2 Factors Affect ng Ve c e Count ng7.3 Counting Accuracy and Quality Assurance..............................................................................................

7.4 Durat on of Count ng

7.5 Count ng Proce ures ................................................................................................................................

.5.1 Intersection

.5.2 Stra g t Roa s

.5.3 rban Roads ................................................................................................................................

.5.4 Rura Roa s 6

.5.5 Dual Carriageways .......................................................................................................................


7/54


Roads Department

TYPI AL ONVER ION OF TRAFFI OUNT .................................................................................3

.1 enera .....................................................................................................................................................

.1.1 Conversion of Average Daily Traffic to Annual Average Daily Traffic 7

.1.2 Conversion of Peak Hour Traffic to Average Daily Traffic (ADT ..............................................

.2 Reference Curves an S ort-Term Counts

9 ANALYSIS AND PR SENTATION OF TRAFFIC DATA........................................................................ 0

9.1 Genera 4

9.2 Layout of Ana ys s ....................................................................................................................................

.2.1 Computer Analysis 4

.2.2 Manua Ana ys s

9.3 Data Entry and Analysis ...........................................................................................................................

.3.1 enera

.3.2 Data Entry .....................................................................................................................................

.3.3 Data Analysis

9.4 Report ng

9.5 Presentation ..............................................................................................................................................4

REFERENCES .......................................................................................................................................................4

APPENDICES

1 Monthly Traffic Census Count Survey Form ................................................................................................... 7

2 Roa Traf c Census Count Survey Form

3 Traffic Census Survey Form ............................................................................................................................4

1 Summary Traf c Data for Trans Kga aga H g way - 1990 5

2 Summary Traffic Data for Trans Kgalagadi Highway 2000 ......................................................................... 1

: Abbrev at ons

LIST OF TABLES

ab e 2.3.1 Summary of recommen e traf c ow eve s 1

able 3.3.1 Traffic count classification for axle load requirements.............................................................

ab e 3.3.2 WIM p ot pro ect n Botswana

able 4.1.1 Percentage traffic flow fluctuation by direction Station N9 2

ab e 8.1.1 Typ ca traf c convers on factors..............................................................................................

able 9.2.1 Analysed manual traffic count data 42

ab e 9.5.1 Summary of AADT for fferent ve c e c asses ......................................................................

able 9.5.2 AADT by direction of flow

LIST OF FIGURES

gure . . e ma n nat ona roa networ

gure 3.3.1 Typ ca ax e con gurat on n t e country .................................................................................igure 4.1.1 Peak traffic flow variation 22

gure 4.1.2 Da y traf c ow var at on........................................................................................................

igure 4.1.3 Annual traffic flow variation by direction.................................................................................2

gure 4.1.4 Typ ca annua traf c growt var at ons

gure . . typ ca con gurat on o an automat c tra c count ng networ ............................................2

igure 5.3.1 Layout of traffic loops on the road surface

gure 5.3.2 A typ ca equ pment for traf c ata co ect on .........................................................................2

igure 7.2.1 Different average daily traffic flow levels

gure 8.2.1 Reference curve

igure 9.5.1 Typical traffic flow variations. ..................................................................................................4


8/54


Roads Department Roads Department


9/54

9Traffic Data Collection and Analysis Chapter 1Introduction

Roads Department

1 INTRODUCTION

. Backgrounotswana w t a popu at on of 1.70 m on n t e year 2001 s pro ecte

to have a population of 2.25 million in 2016 and a projected annual traffic

growt of 8%. W t t s scenar o, t s expecte t at t e tota number o

vehicles using the Public Highway Network (PHN) will increase substan-

t a y. T e resu t ng ve c e eet, s expecte to e over 300 ve c es per

,000 persons in the year 2016 from the current level (2002) of approxi-

mately 90 per 1,000 persons. This indicates a potential demand for invest-

ment n transport nfrastructure. Proper ut sat on of suc uge nvest-

ments necessitates systematic planning for need-based development. Such

nee - ase eve opments nc u e eterm nat on of t e requ re capac ty

expansion, provision of additional road infrastructure, improvement o

ex st ng roa s, pr or t sat on of fferent eve opment p ases an fore-casting of which is possible upon collection of traffic data. This is done in

or er to e m nate bott enec s n bot nternat ona an oca nter-urban

roa transport towar s prov ng an ef c ent an effect ve roa transport

sys em.

he concept of forecasting the future use of the road network in terms o

traf c oa ng an ow, s genera y an accepte approac wor -w e.

he techniques used have become almost standard in both developing

an eve ope countr es. T e accuracy of traf c ata co ect on an t e

subsequent predictions are of paramount importance in the fulfilment o

an appropr ate p ann ng, es gn, ma ntenance mon tor ng an manage-

ment of t e roa networ .

n t e past, rout ne co ect on of traf c ata n t s country was not

considered important for the development and management of the roa

networ . In t e ear y 1970s t was rea se t at a w e var ety of nfor-

mation is required in respect of traffic characteristics for proper mainte-

nance, p ann ng, es gn, ma ntenance an management of t e nat ona

road network. This realisation emanated from concerns raised with regar

to t e amount of traf c vo ume , t e compos t on of t e fferent types

of vehicles, their speed, total gross weight, number of axles, axle loads

and origin and destination of the journeys. Most of this information result

n assessment of progress ve or rap eter orat on of t e roa networ

towards estimating additional cost required to sustain it.

s a result, attempts are now being made to adopt suitable road traffic

met o o og es for con uct ng roa traf c surveys, w c are bot tec n -

cally and scientifically sound, and operationally convenient to execute

un er t e countrys preva ng con t ons. T s nc u es t e use of bot

manual and automatic traffic counters, together with computer analysis o

t e co ecte traf c ata. Dur ng t e p ann ng, es gn, construct on an

maintenance period of the road network, traffic data becomes an essential

e ement n ec s on-ma ng, an t erefore t e format an t e accuracy o

ata collection and analysis is critical. It is with this view that this guide-

line on traffic data collection and analysis has been prepared.


10/54

10 Traffic Data Collection and AnalysisChapter 1Introduction

Roads Department

Figure 1.1.1: The main national road network.

1.2 Purpose and Scope of the Guidelines

n order to facilitate the assessment of present and future traffic demands,

for t e eve opment of nee -base nfrastructure accurate nformat on

and continuous monitoring of traffic by appropriate methods is necessary.

mp ement ng aut or t es must t erefore ensure t at suf c ent an appro-

priate data is available to undertake necessary planning, design, construc-

t on an ma ntenance of t e countrys roa networ , w c s a me

at meeting the prevailing traffic flow, future traffic growth and loading

without considerable deterioration in the quality of service.

his guideline has therefore been prepared with the main aim being to

prov e bas c nformat on, concept an pr nc p es w t respect to traf c

data collection and analysis. There are various methods of data collection

ava a e an use y fferent organ sat ons nst tut ons. T s gu e ne,

therefore, is only intended to provide guidance in respect of data collec-

t on an ana ys s, an a ows for var at on n t e met o o og es a opte

by different users, planners, developers, funding authorities, etc.

e ene c ar es of t s gu e ne are Roa s Department, ot er

Ministries/Departments, local authorities, educational institutions, the

pr vate sector an n v ua s.


11/54

11Traffic Data Collection and Analysis Chapter 1Introduction

Roads Department

1.3 Structure of the Guidelinehe guideline comprises of nine Chapters and six Appendices.

Chapter 1.0

his chapter gives a broad background on the concept of traffic data col-

ect on. T e c apter a so nc u es scuss on of t e purpose an scope o

the guideline.

Chapter 2.0

rov es t e ro e an funct on of t e gu e nes w t respect to traf c ata

co ect on an ana ys s, t e types of traf c counts, genera spec cat ons

and quality assurance of data being collected.

Chapter 3.0

scuss an g ve extens ve gu ance on traf c ata co ect on n Botswana

he chapter starts with an overview of the current practice in Botswanaan prov es essent a approaches requ re to con uct a proper traf c ow/

volume survey. This Chapter also discusses a process or methodology for

se ect on of count ng ocat ons, ve c e c ass cat on an con gurat on,

w c may vary epen ng on t e resu t ng use of ata.

hapter 4.

rov es scenar os for count ng c aracter st cs n respect of c o ce o

count ng system, traf c vo ume c aracter st cs an frequency of traf c

counts. It also provides guidance regarding data collection and storage.

Chapter 5.0

Cons ers resources requ re to un erta e t e exerc se, w t respect tostaff composition, qualifications and equipment requirements. As the

wor s to e carr e out a ong motor ze nat ona roa networ , traf c

safety on site is key to the resources required.

Chapter 6.0

his chapter deals with issues of safety being paramount to the whole

exerc se of traf c ata co ect on w t part cu ar reference to s g t ng o

counting sites, provision of road signs and markings, etc.

Chapter 7.0

s s t e core c apter n t s gu e ne, as t prov es gu ance on

vehicle counting, with particular reference to factors affecting vehiclecount ng, accuracy, urat on of counts, count ng proce ures, typ ca

counts conversions an reference curves.

hapter 8.

his chapter presents typical counts conversions and reference curves.

t a so out nes ayout of t e ana ys s, ata entry, ana ys s, an report ng.

Chapter 9.0

his chapter outlines the recommended layout of traffic data analysis,

ata entry an ts presentat on, owever, t e recomme e met o s are

not exhaustive.


12/54

12 Traffic Data Collection and AnalysisChapter 2Traffic Data Collection

Roads Department

2 TRAFFIC DATA

COLLECTION2.1 Role and Function

raf c Data Co ect on an pro ect ons t ereof of traf c vo umes are

basic requirements for planning of road development and management

sc emes. Traf c Data forms an ntegra part n t e sc ence of escr p-

tive national economics and such knowledge is essential in drawing up a

rational transport policy for movement of passengers and goods by both

government an t e pr vate sectors.

s Gu e ne cons ers t e fact t at traf c ow ata s mportant n

planning of a particular section of the road network and for its subsequent

ma ntenance. Traf c ow pattern appears to be ran om n str but on, as t

reflects peoples motivation in terms of different composition of vehicles

on fferent types of roa s un er vary ng env ronmenta con t ons.

t fo owst en t at ata be ng co ecte s a met o o og ca s tat st cs,

because traffic flow pattern follows a random distribution. Despite such

comp ex t es, t oes fo ow fa r y an c ear y e ne patterns t at are

possible to classify and analyse. Thus, traffic data collection and analysis

fo ows vary ng tren s an p ays an mportant ro e n t e eva uat on an

management of road network schemes.

While taking cognisance of the above, traffic flow data is needed for dif-

ferent purposes y fferent M n str es an or Organ sat ons n Botswana.

e ma or areas for w c t s ata s requ re are:

P ann ng pr or t sat on an pro ect n t at on.

Project design.

P ann ng ma ntenance.

National Transport Statistics.

Roa Safety Measures.

Traffic Control.

ome of the key areas in which traffic flow data is needed for develop-

ment and management of the road network include:

a) Determination of a programme of road widening needs and generalmprovement or strengt en ng of ex st ng roa t roug a programme

of reconstruction and construction of a new roads;

To check the efficiency of the road network by comparing current

traf c vo ume w t t e eve of serv ce or t e ca cu ate capac ty;

c To estab s t e re at ons p between traf c vo ume, number of acc -

dents and causes thereof, as well as determination of the probable

u

To p an pr or t sat on of roa s mprovement sc emes;

Project identification or inception.

Project appraisal and feasibility.

Project screening and planning.

Project Design.

Project Implementation.

Operation, Monitoring and

Maintenance.

Tra c flow data is important in

p anning of a particu ar section

of the road.


13/54

13Traffic Data Collection and Analysis Chapter 2Traffic Data Collection

Roads Department

e To assess econom c bene ts ar s ng from roa s mprovements;

f nvest gat on of var ous capac ty an es gn pro ems for ot roa s

and bridges and parking facilities.

g) Design and improvement of new/existing unctions;

) ss stance n p ann ng new eve opments suc as roa s n a new

town, subdivisions, land use, which generally includes shopping

centres, ote s, commerc a an n ustr a comp exes, serv ce stat ons

and other traffic generators activities;

i) Determination of warrants or the need for implementation of traffic m-

provement an traf c contro measures, suc as sync ron se /coor -

nate traf c s gna s, stop s gns, one way roa s, no entry, etc;

To stu y future traf c tren s an ass st ng n pre ct ng traf c owsin the future for a given period;

k) To classify roads on their functional basis.

n addition to the above the following are typical specific needs:

a Assessment of pavement performance t roug traf c surveys an

Period monitoring of selected sections;

b) Ascertaining appropriate/optimal timings for maintenance interven-

t ons an re ab tat on nee s of var ous roa s countryw e;

c Esta s econom c an soc a mp cat ons of es gn an feas ty

studies of all development projects countrywide;

Estab s t e use of t e roa networ by ve c es of fferent catego-

ries, traffic distribution, etc.

2.2 Types of Traffic Countst s essent a to know the magn tude of traf c data requ red or to be co ect-

ed, which will then determine its quality and type of vehicle classification to

be adopted. Traffic counting falls in two main categories, namely; manual

counts an automat c counts. T ere s no st nct fference etween t e twomethods however, the economic use or selection of an appropriate method

of traf c count ng s a funct on of the eve of traf c ow and the requ red

data quality. This difference can be deduced from the discussions of the

respect vemethods be ow, and n the subsequent chapters.

2.2.1 Manual Countshe most common method of collecting traffic flow data is the manual

met o , w c cons st of ass gn ng a person to recor traf c as t passes.

his method of data collection can be expensive in terms of manpower,

ut t s nonet e ess necessary n most cases w ere ve c es are to e

classified with a number of movements recorded separately, such as at

ntersections.

Manual traffic counting in progress.


14/54


Roads Department

About 64 permanent manual traffic-counting sta-

tions have been established countrywide by Roads

Department.

Security for counting equipment is veryessential.

t ntersect on s tes, t e traf c on eac arm s ou be counte an

recorded separately for each movement. It is of paramount importance

t at traf c on roa s w t more t an one ane are counte an c ass e y

irection of traffic flow.

ermanent traffic-counting teams are normally set up to carry out the

counting at the various locations throughout the road network at set

nterva . T e urat on of t e count s eterm ne pr or to commencement

of traffic counting and it is dictated by the end use of data. The teams

are manage an superv se by t e tec n ca staff to ensure ef c ent an

proper collection of data. (Survey Forms used for manual traffic-counting,

s g ven nAppendix A1, A2 an A3.

.2.2 Automatic Countse etect on of ve cu ar presence an roa occupanc es as stor ca y

been performed primarily on or near the surface of the road. The exploita-

t on of new e ectromagnet c spectra an w re ess commun cat on me an recent year, has allowed traffic detection to occur in a non-intrusive

fas on, at ocat ons above or to t e s e of t e roa way. Pavement-base

traffic detection currently relatively inexpensive, will be met with fierce

compet t on n t e com ng years from etectors t at are berate from t e

roa surface.

he most commonly used detector types are:

) Pneumatic tubes.

ese are tubes p ace on t e top of roa surfaces at ocat ons w ere

traf c count ng s requ re . As ve c es pass over t e tu e, t e resu t ng

compression sends a burst of air to an air switch, which can be installen any type of traf c count ng ev ces. A r sw tc es can prov e accurate

axle counts even when compressions occur more than 30 m from the traffic

counter. A t oug t e fe of t e pneumat c tubes s traf c epen ant as

they directly drive over it, it is used worldwide for speed measurement

an ve c e c ass cat on for any eve of traf c. Care s ou be exerc se

n placing and operating the system, to ensure its efficient operation an

m n m se any potent a error n t e ata.

ii) Inductive loops.

nductive loop detector consists of embedded turned wire from which it

gets ts name. It nc u es an osc ator, an a cab e, w c a ows s gna s

to pass from the loop to the traffic counting device. The counting devices activated by the change in the magnetic field when a vehicle passes

over t e oop. In uct ve oops are c eap, a most ma ntenance-free an

are currently the most widely used equipment for vehicle counting an

etect on. S ng e oops are ncapab e of measur ng ve cu ar spee an

the length of a vehicle. This requires the use of a pair of loops to estimate

spee by ana ys ng t e t me t ta es a ve c e to pass t roug t e oops

nstalled in series. An inductive loop can also, to a certain degree, be use

to etect t e c ass s e g ts an est mate t e number of ax es.

y us ng t e n uct ve oops, t e engt of t e ve c e s t erefore er ve

from the time taken by the vehicle to drive from the first to the secon


15/54

15Traffic Data Collection and Analysis Chapter 2Traffic Data Collection

Roads Department

Automatic counting station secured and oper-ated by solar power.

Typical installation of inductive loops.

A weigh in-motion speed and axle detectionpad.

Weigh-in-Motion along Trans-KgalagadiHighway.

oop r v ng t me an t e t me ur ng w c t e ve c e was over t e

rst and the second loop (cover time). The resulting length is called the

e ectr ca engt , an s n genera ess t an t e actua engt of t e pass-

ng vehicle. This is caused by the built in detector threshold, the roa

surface mater a , t e fee er engt , t e stance between t e bottom of t e

vehicle and the loop, but also, to a large extent, the synthetic materials

used in modern cars. The system could be used for any level of traffic.

iii) Weigh-in-Motion Sensor types.

var ety of traf c sensors an oops are use wor -w e to count, we g

and classify vehicles while in motion, and these are collectively known

as We g In Mot on WIM sensor systems. W ereas sensor pa s can be

used on their own traffic speed and axle weighing equipment, they are

tr gge by ea ng n uct ve oops p ace before t em on t e roa -

be . T s scenar o s a opte w ere ax es, spee an stat st ca ata are

required. Some notable traffic sensors are:

Bending Plates hich contains strain gauges that weigh the axles of

pass ng veh c es. Cont nuous e ectr c s gna s are sent to the stra n gauges,

and these signals are altered as the plates are deflected by dynamic

ve cu ar we ght and measure the ax e o the pass ng veh c es.

Capacitive Strip s a th n and ong extruded meta used to detect pass-

ng ax es. The force of vert ca pressure app ed to th s str p by a whee

alters its capacitance, which can be converted to a wheel-weight measure

when re ated to the speed of the veh c e. Capac t ve str ps can be used

for both statistical data and axle configuration.

Capacitive Matfunctions in a similar manner as the capacitive striput t s es gne to e mo e an use on a temporary as s on y.

P ezo-electr c able s a sens ng str p of a meta c cab e that responds to

vertical loading from vehicle wheels passing over it by producing a

correspond ng vo tage. The cab e s very good for speed measurement

and axle-space registration, and is relatively cheap and maintenance free

ke a nduct ve oop f nsta ed correct y.

iv) Micro-millimetre wave Radar detectors.

Radar detectors act ve y em ts rad o act ve s gna s at frequenc es rang ng

rom the ultra-high frequencies (UHF) of 100 MHz, to 100 GHz, and can

reg ster veh cu ar presence and speed depend ng upon s gna s returned uponreflection from the vehicle. They are also used to determine vehicular vol-

umes and c ass cat ons n both traf c d rect ons.

a ar etectors are very tt e suscept e to a verse weat er con t ons,

and can operate day and night. However, they require comparatively high

levels of computing power to analyse the quality of signals.

v) Video Camera.

ideo image processing system utilise machine vision technology to

etect vehicles and capture details about individual vehicles when necess-

ary. A v eo process ng system usua y mon tors mu t p e anes s mu ta-


16/54


Roads Department

R ad D ar men

neous y, an t erefore t requ res g eve of comput ng power. Typ -

cally, the operator can interactively set the desired traffic detection points

anyw ere w t n t e systems v ew area.

lgorithms are used to extract data required for the detection of the raw

ata fee s. Due to t e comp ex ty of t e mages, t s not recommen e

t at t ey s ou be processe out oors as t s can g ve poor resu ts. T e

system is useful for traffic counting and give a +/- 3% tolerance, and is

not appropr ate for ve cu ar spee an t e r c ass cat on.

2.3 General Specificationseneral specification provide guidance for the allocation and distribution

of traffic counting points along the national road network. This is done in

or er to ma nta n cons stency an atta nment of appropr ate compar sons

of data between counting stations.

rom traffic counts observations at different locations in the country, it

as een not ce t at manua counts may e prone to ncrease uman

rr w en t e number o enumerators are ncrease commensurate w t

ncreas ng traf c ow.

n the context of this guideline, temporary automatic counting can be to

a m n mum of seven- ay cont nuous counts four t mes a year. At t ese

stations both the loops and the roadside cabinet have to be installed. The

equ pment for automat c count ng can t erefore e move aroun to t e

various temporary stations as required. Some of the temporary stations

can ater be converte to permanent stat ons, as requ re , w en t e traf c

flow increases.

umber re erencing o automatic countingstat ons.

Proper maintenance of counting stations ensuresminimum disturbance to the system and reducespotential damage and vandalism.

A national dual carriageway road with a con-tinuous traffic flow in an urban environment.

lassi cation o vehicles by type re ects theix in the traffic stream passing the counting

point.

Number ofVehicles(paved)

Number ofVehicles

(unpaved)

Automatic Counts Manual Counts

> 2000 Required at permanentstations continuouslythroughout the year.

1000 - 2000For specific period tobe adopted (temporaryautomatic counting).

Could be used whenappropriate, that isepending on the levelf traffic flow.

< 1000Counted at intervalsfor seven consecutivedays, 16 hours per day,four times a year.

> 50

Counts for sevenconsecutive days for 16hours once a year or asdeemed appropriate forthe end use of data.

0 - 50

Counts for sevenconsecutive days, 16hours a day, once everyyear or as deemedappropriate.

< 20

N traffic counts shouldbe carried out unlessrequired for specialreasons.

Table 2.3.1: Summary of recommended traffic flow counting levels.


17/54

1717Traffic Data Collection and Analysis Chapter 3Chapter 3Traffic Data Collection, Practises in Botswana

Roads Department

R a Dep rt e

3 TRAFFIC DATACOLLECTION, PRACTISES

IN BOTSWANA

1.1 Generalhe primary source of traffic data collection in Botswana is through

estab s ment of regu ar manua traf c count ng programmes an spon-

taneous automatic counters along the public highway network. Presently

(2002), 64 permanent manual traffic count stations are located country-

w e manage by Roa s Department. In a t on, t ere are ot er stat ons

established for special traffic counts such as ad-hoc (short-term) counts or

spec a counts, w c are ma n y con ucte on roa s w t out permanent

stations. Origin Destinations survey is a special way of carrying outtraf c counts survey, w ere y t e ata co ecte re ates to t e use of t e

road by vehicle category.

e var ous types an met o s use to co ect traf c ata not on y pro-

vide a good and valuable coverage of the required traffic information for

ec s on ma ng an p ann ng of bot eve opment an ma ntenance o

the national road network. On this basis, all local authorities, institutions/

organ sat ons, etc., are urge to use t s Gu e ne as a too n t e mp e-

mentation of a traffic counting system to enable creation of a national

ata base at t e r respect ve eve s of operat on for a c asses of roa s

ncluding tertiary roads, access roads, town/city roads, village roads, an

many ot ers not accounte for n t s Gu e ne for proper management

of t e roa nfrastructure.

3.2 Selection of Counting Sitess stated earlier, the typical traffic counting system used by Roads

epartment compose of 64 permanent manua traf c count ng stat ons

an a t ona var ous spec a count ng stat ons.

spec c ocat on for count ng s te permanent or temporary must be

etermined on site. Where automatic counting system is to be used, the

exact ocat ons of oops s ou be ec e w e ta ng cogn sance o

the potential use of data collected. The following should be kept in min

efore ec ng on t e count ng s te:

T e roa sect on s ou ave un form geometr c c aracter st cs a ong

t e roa engt an be away from unct ons;

Location should be on a horizontal (flat) and geometrically straight

roa sect on;

Section of the road to have an uninterrupted traffic flow;

Sect ons w ere te ep one nes or ra o mob e are eas y

access b e or can be nsta e , f poss b e;

Section to have very little pedestrian or animal traffic;

Sect on to meet safety requ rements.

A typical traffic counting station.A typical traffic counting station.

A typical straight road.A typical straight road.


18/54

1818 Traffic Data Collection and AnalysisChapter 3Chapter 3Traffic Data Collection, Practises in Botswana

Roads Department

n areas w ere trun roa s are n t e v c n ty of ma or c ty m ts or town

boundaries (such as Gaborone, Francistown and Lobatse), the locations o

count ng s tes us ng oops s ou be oo e nto n more eta . In uc-

tive loops at traffic signals may also be used for traffic counting in ma or

cities and towns.

o ensure that adequate attention is given to sections of roads constitut-

ng a spec c traf c ow, roa s s ou be v e nto un form sect ons

accor ng to traf c c aracter st cs.

3.3 Vehicle Classification andConfiguration

lthough there are various classification approaches available to various

nst tut ons/organ sat ons or en users of t s Gu e ne, t e respons b e

organ sat on/ nst tut on may use w c ever met o s eeme appropr ate

and applicable for the purpose at hand, depending on the quality of data

requ re .

Manual Count classification

Manual traffic flow count is categorised by a visual assessment of the

ve c e s ze an con gurat on of ax es. T e current manua traf c ow

data collection system in Botswana classifies vehicles into nine categories

as follows:

Passenger ars .............................................................................. (1M)

Pick up or Van less than 1.5 tons (2M)

Trucks, more than 1.5 and less than 5 tons (3M)

Light Lorries, more than 5 and less than 10 tons .......................... (4M) Medium Lorries, more than 10 and less than 20 tons (5M)

Heavy Lorries, more than 20 tons................................................. (6M)

Heavy Busses (7M)

Mini Busses (8M)

Agriculture Tractors...................................................................... (9M)

ruc s- two ax es, 4 x 4 ve c es of 1.5 tonnes but ess t an 5

tonnes.

Light Lorries- commercial vehicles of two axles. These are vehicles

w t 5 tonnes but ess t an 10 tonnes.

Medium Lorries- vehicles with 3 or 4 axles and having weight o

more t an 10 tons but ess t an 20 tons.

eavy Lorries - all lorries with 5 axles or more and weight more than

tons.

It is important to know the most damagingIt is important to know the most damagingategories of vehicles at the earliest stage ofategories of vehicles at the earliest stage of

the design.the design.

Load distribution over the various vehicle axleLoad distribution over the various vehicle axleonfigurations determines the vertical load ononfigurations determines the vertical load on

the road.the road.


19/54


Roads Department

Table 3.3.1: Traffic count classification for axle load analysis.

W ereas t e above are t e ve c e categor es n use, t s mportant to

note that traffic counting and classification is a dynamic process an

ts pro ect base . Tab e 3.3.1 above s a typ ca c ass cat on of eavy

vehicles by their configuration for purposes of axle load analysis. In this

context, nformat on e ng co ecte s ou ensure mu t p e uses of ata

co ecte e t er manua y or by automat c system.

Automatic Traffic Countsfully equipped Automatic Traffic Counters with inductive loops an

WIM-sensors we g pa as a potent a of c ass fy ng traf c as requ re

by means of the following classification parameters:

- Number of ax es.

- We g t of eac ax e.

- Axle spacing.

- Spee .

- Vehicle length.

- C ass s e g t.

Where on y nduct ve oops are used the parameters that can be reg stered are:

- Speed.

- Ve c e engt .

- To a certa n egree C ass s e g t.

- The number of axles can be estimated.

s a result, it is appropriate to use automatic traffic counters with either:

- In uct ve oops on y w ere traf c ow ata s requ re ; or

- Inductive loops and WIM sensors where both traffic flow data and

ax e or spee measurements are requ re .

Notwithstanding the above, where only inductive loops are used it is possible

to count 5 classes using the following axle spacing classification:

1L L g t ve c es, Cars, Tax s, Vans an P c -Ups engt 0 - 5.5 m

2L Unarticulated trucks, 2 - 3 axle Trucks, Buses ............. length 5.6 - 7.5 m

3L Unart cu ate us, Buses an Truc s engt 7.6-12.4 m

L Heavy trucks and very ong Buses ............................ ength 12.5 - 15.9 m

5L Heavy trucks (length 16.0 22.0 m

Vehicle classification can be determined by theVehicle classification can be determined by thenumber of axles and the gross vehicle mass.number of axles and the gross vehicle mass.

Ensure that traffic data collected has a mul-Ensure that traffic data collected has a mul-tiple use and captures relevant classes or typestiple use and captures relevant classes or types

of vehicles for the inten e purpose.of vehicles for the inten e purpose.

Heavy vehicle category Definitions

Buses Seating capacity of 40, or more

Medium Goods vehicleMGV

- 2 axles, incl. steering axle, and- 3 tonnes empty weigh, or more

Heavy Goods VehicleHGV

- 3 axles, incl. steering axle and- 3 tonnes empty weigh, or more


20/54

2020 Traffic Data Collection and AnalysisChapter 3Chapter 3Traffic Data Collection, Practises in Botswana

Roads Department

ith inductive loops and WIM censors installed it is possible to count 7

c asses as fo ows:

1W Car, Taxi and Vans......................... (2 axle, length 0-5.5 m)

2W 4W-dr ve P ck-Ups same as 1W, but w th h gher chass sW 2-axle Trucks and Buses................ (length 5.6 7.5 m, weight - 2 tons

4 3-ax e Trucks and Buses.............. ength 5.6 - 7 m, we ght 2 - 20? ton

5W 2-3 -axle Buses and Trucks ........... (length 7.6 - 12.4 m

6 4-5-ax e Trucks .............................. ength from 12.5 m

W 6 or more ax e Truc s engt from 16.0 m

Where counts have been done w th both WIM and nduct ve oops t s often

f cu t to spec fy the c ass cat on of aggregated data. A better so ut on s to

store t e fo ow ng measure ata for eac ve c e type:

Date and time. Speed. Total length .

A ax e spac ng. A ax e we ghts.

hen data has been stored for each veh c e, var ous opt ons can be se ected

an t e computer can e programe to perform var ous ca cu at ons asmay be requ re .

ith the use of inductive loops in each lane it is possible to measure thevehiclespeed very accurately. Most counting equipments have speed regi-strat on wh ch then group veh c es nto speed categor es, aga n these can be

used to group the d fferent veh c es by the r ax e spac ng. Some count ng

equ pment as t e fac t es to reg ster t e average spee an var at on of t e

speed.

or eac rect on t e fo ow ng ve c e spee ata s ou e co ecteourly:

Average speed of all vehicles Speed variation of all vehicles 85 faster than 15 faster t an e c es > 60 m/ e c es > 80 m e c es > 100 m/ ehicles > 120 km/h ehicles > 140 km/h

Number of vehicles < 2 sec behind vehicles ahea Average speed length for group 1L Average speed length for group 2L Average spee engt for group 3L Average spee engt for group 4L Average spee engt for group 5L

In recent years substantial efforts have been made by various traffic authori-

ties all over the world to develop/use suitable Weigh-in-Motion (WIM)

equipment, in some cases these efforts have not been fully successful.

istory has shown that the use of automated systems by Roads Depart-ment has not been too success u or the co ect on o tra c data us ng both

WIM and statistical traffic data. The main problems are associated with the

A typical bus withA typical bus with 2 axles.2 axles.

Cars, vans and pick-ups constitute high pro-Cars, vans and pick-ups constitute high pro-portion o tra c in t e country.portion o tra c in t e country.

A speed, axle and traffic flow data collectionA speed, axle and traffic flow data collection

station with a lockable counting equipment box.station with a lockable counting equipment box.


21/54


Roads Department

Single tyres on front and rear axles

Single tyres on front axleTwin tyres on rear axle

Single tyres on front axle

Twin tyres on rear axlesTwo rear axles

Single tyres on front axleTwin tyres on rear pair of axles

Two rear axles

Single tyres on front pair of axlesSingle tyres on rear pair of axles

Single tyres on front pair of axles

Twin tyres on rear axle

Single tyres on front pair axlesTwin tyres on rear pair of axles

TRAILERSSingle tyres on both axles

Single tyres on front axle

Twin tyres on rear axle

Twin tyres on both axles

11

12

1.11

1.22

11.11

11.2

11.22

1.2 + 1.1

1.2 + 1.2

1.2 + 2.2

1.1 - 1

1.1 - 11

1.1 - 22

1.2 - 1

1.2 - 11

1.2 - 2

1.2 - 22

1.22 - 2

1.22 - 22

1.22 - 111

1.22 - 222

Single tyres both axles of tractor

Single tyres axle of trailer

Single tyres on both axles of tractorSingle tyres on both axles of trailer

Single tyres on both axles of tractor

Twin tyres on both axles of trailer

Single tyres on front axle of tractorTwin tyres on rear axle of tractor

Single tyres on axle of trailer

Single tyres on front axle of tractor

Twin tyres on rear axle of tractorSingle tyres on both axleS of trailer


Twin tyres on rear axle of tractor

Twin tyres on axle of trailer

Single tyres on front axle of tractorTwin tyres on rear axle of tractor

Twin tyres on both axle of trailer


Twin tyres on both rear axles of tractorTwin tyres on rear axleS of trailer

Single tyres on front axle of tractorTwin tyres on both rear axles of tractor

Twin tyres on both rear axles of trailer

Single tyres on front axle of tractorTwin tyres on both rear axles of tractor

Single/twin tyres on axles of tractor

RIGID - CHASSIS COMMERCIAL VEHICLES ARTICULATED COMMERCIAL VEHICLES

Lack of proper maintenance can lead to someLack of proper maintenance can lead to somestations being rendered un-operational andstations being rendered un-operational andloss of important traffic data.loss of important traffic data.

A damaged road surface exposing inductionA damaged road surface exposing inductioncables is prone to damage.cables is prone to damage.

igure . . : ypica ax es con gurations in t e country.

requent breakdown of the equipment. Firstly, this has been due to lack o

adequate or susta nab e ma ntenance. In the future, when purchas ng new

traffic counting equipment, the maintenance of the equipment should be

nc u e as part of t e supp y agreement. T s nc u es per o c, emergency

and routine maintenance activities.

ollection of axle loads data could therefore be dealt with as a separate task

us ng mob e we ghbr dges, f the we ght measurements are separated from

the traffic flow ata collection.

A pilot project in 1997/98 involving both Truvelo and Micros equipmentA pilot project in 1997/98 involving both Truvelo and Micros equipment waswas

carried out in Botswana by Roads Department. In brief, the conclusion were:carried out in Botswana by Roads Department. In brief, the conclusion were:

Both systems can work satisfactorily only under supervision of highlyBoth systems can work satisfactorily only under supervision of highlyspecialised personnel. Maintenance and calibration should be done atspecialised personnel. Maintenance and calibration should be done atregular intervals.regular intervals.

Moderate changes in processing software, calibration algorithms andModerate changes in processing software, calibration algorithms andsystem refinement should always be a continuous process as the tech-system refinement should always be a continuous process as the tech-

nology develops.nology develops.

It is virtually impossible to operate the system without close interactionIt is virtually impossible to operate the system without close interactionand assistance of the manufacturers.and assistance of the manufacturers.

From this pilot project the Mikros System equipment was found to be havingFrom this pilot project the Mikros System equipment was found to be having

less operational constraints, and there by being extensively used in Botswana.less operational constraints, and there by being extensively used in Botswana.

Table 3.3.2: WIM pilot project in Botswana.


22/54

22 Traffic Data Collection and AnalysisChapter 4Counting Characteristics

Roads Department

4 COUNTINGCHARACTERISTICS

.1 Traffic Flow

4.1.1 Frequency o Tra c ounts

n or er to pre ct traf c ow vo umes t at can be expecte on t e roa

network during specific periods, cognisance should be taken of the fact

t at traf c vo umes c anges cons erab y at eac po nt n t me. T ere are

three cyclical variations that are of particular interest:

Hourly pattern: T e way traf c ow c aracter st cs var es t rougout the day and night;

aily Pattern: T e ay-to- ay var at on t roug out t e wee ; an

Monthly and yearly Pattern:The season-to-season variation

t roug out t e year.

When analysing the traffic one must also be aware of the directional dis-

tr but on of traf c an t e manner n w c ts compos t on var es.

Hourly patterns

Typical hourly patterns of traffic flow, particularly in urban areas, generally

s ow a num er of st ngu s a e pea s. Pea n t e morn ng fo owe y a

lean flow until another peak in the middle of the afternoon, after which there

may be a new peak n the ate even ng. The peak n the morn ng s o ten

more sharp by reach ng the peak over a short durat on and mmed ate y

dropping to its lowest point. The afternoon peak on the other hand is

c aracter se y a genera y w er pea . T e pea s reac e an sperse

over a longer period than the morning peak. However, in urban satellite

towns, the morn ng peak may be too ear y and even ng peak may be too ate

in comparison to the principal towns without significant midday peak.

Figure 4.1.1: Traffic peak flow variation.

Exper ence from many countr es s ows t at a t oug traf c vo umes may

grow over time, the relative variations of traffic at the various hours of the

ay of a mont are often qu te cons stent year after year.

0

200

400

600

800

1000

1200

Time in hours

0

200

400

600

800

1000

1200

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

An earth/sand road in a rural area.


23/54

23Traffic Data Collection and Analysis hapter 4Counting Characteristics

Roads Department

Daily patterns

he traffic volume generally varies throughout the week. The traffic dur-

ng t e wor ng ays Mon ay to Fr ay may not vary su stant a y, ut

the traffic volume during the weekend is likely to differ from the working

ays on fferent type of roa s an n fferent rect ons. In Botswana

many of the urban population goes to the rural areas during the weekends,

hence a high variation of traffic on the urban rural connector roads during

wee ays an wee en s.

Figure 4.1.2: Daily traffic flow variations.

he pattern from Monday to Friday is often relatively consistent, a part

from Monday morn ng and Fr day afternoon traf c ow. The pattern dur ng

the weekend may vary considerably. The pattern also varies from Saturdays

to Sun ays. T e pattern ur ng t e wee en s s a so e y to s ow more

seasona var at on than dur ng the work ng days.

Monthly and yearly patterns

he monthly and yearly pattern normally reflects the seasonal variation o

traf c ow. T e pattern may vary for passenger cars an ve c e trans-

porting goods. In Botswana, variation between traffic flow during the wet

season an ur ng t e ry season s ns gn cant.

igure 4.1.3 : Annual traffic flow variations by irection.

now e ge of ow traf c ows var es for fferent ve c e c asses may

be very useful when designing programmes for axle load control, design

of roa s mprovement upgra ng sc emes an sett ng ma ntenance

programmes an pr or t es. A t oug Botswana s not nown to ave

significant seasonal variation in traffic flow, such occurrences should be

Traffic Flow on Different Days of the Week

0

2000

4000

6000

8000

10000

12000

Traffic Flow Counting Stations

Monday

Tuesday

Wednesday

Thursday

Friday

Sartuday

Rakhun

a

LobatseSo

uth

Otse

Gaboron

eSo

uth

GaboroneNo

rth

Rasesa

0

500

1000

1500

2000

2500

3000

Year

AADT

Palapye Direction

Mahalapye Direction

M/Drift Direction

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000


24/54

24 Traffic Data Collection and AnalysisChapter 4Counting Characteristics

Roads Department

cons ere so as to a equate y account for t em t roug app cat on o

appropriate conversion factors.

Monitoring traffic growth variations

he number of motor vehicles license in Botswana has shown an increase

n traf c ow stead y over the years. In the 1980s the annua traf c growth

was 11-12 % an t was g est on pave roa s, w e n 2000 t e growt

rate was in the order of 7 8%.

Table 4.1.1 : Percentage tra c ow uctuations by direction at station N9 .

lthough the annual traffic growth may have been less in the 1990s

t an n t e 1980s, t e present nat ona growt rate of 8% s qu te g

and needs to be monitored to ensure timely provision or improvement ot e nat ona roa networ . It s t erefore cr t ca to mon tor growt rates

on newly paved roads on monthly or quarterly basis immediately after

opening to traffic to guard against rapid growth due to presence of the

surface roa .

Figure 4.1.4: Typical annual traffic growth variations.

W en a grave roa s upgra e to pave stan ar , t e traf c pattern on

both the road concerned and on the network changes due to the improve

state of t e roa . Some of t e o pave roa s may ave a re uce traf c

flow and also an overall change in the composition of vehicles using the

A well maintained gravel secondary road tomeet traffic demand.

Palapye Mahalapye M/Drift Total

Direction Direction Direction

1990 19 20 21 60

1991 3 6 -4 5

1992 7 17 23 7

1993 6 -21 14 -1

1994 -1 25 -3 1

1995 24 22 10 56

1996 -7 -10 -3 -20

1997 4 10 18

1998 21 20 39 80

1999 3 2 21 2

2000 4 8 18 0

pgrading of a road is a function of bothtraffic demand(growth) and prevailing humanpopulation to be served.

0

1000

2000

3000

4000

5000

6000

Year

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000


25/54

25Traffic Data Collection and Analysis hapter 4Counting Characteristics

Roads Department

While traffic counting mainly aims at motorisedtraffic, animal traffic are of importance at designstages for etermination of appropriate ansafe crossing points.

road. Th s s often a resu t of traf c red str but on and routes re-pr or t sat on

by different road users.

4.1.2 Data Acquisition

ost counting systems are used for daily, weekly or monthly administra-

t ve count ng per o s, at an nterva of one our or as requ re by t e en

user. T ese a m n strat ve counts are carr e out cont nuous y, un ess t s

moved to a new counting location. The traffic data at the end of a counting

per o s remote y broug t to a centra po nt for process ng an subse-

quent storage. Where communication lines are not provided to retrieve

ata remote y a aptop computer s use to retr eve ata from s te t en

own loaded into a PC for subsequent analysis.

Figure 4.1.5: A typical configuration of an automatic traffic counting network.

n certa n s tuat ons more eta e nformat on of traf c ow may be

needed. In such cases the counting equipment is set for shorter counting

nterva s genera y between 5 to 15 m nutes.

t s mportant t at t e ro e -out count ng ata ata from numerous

counts and dating back some years) can be identified unambiguously

rom ts storage. Part of t s protect on s ou be foun n t e process ng

software and the rolling-out procedure should explicitly be stated in the

systems operating manual.

t is recommended that the rolling-out program uses the number of the

census s te an t e ate of t e start of t e a m n strat ve count ng per o ,to create a unique file name, which allows files to be identified at ease.

herefore it is recommended that the following information is stored:

Name of Aut or ty/Organ sat on.

Version number of rolling-out software.

Version number of counting software.

Counter ent cat on number.

Location code.

Locat on name.

Lane ent cat on.

Counter configuration

Visual

Display


26/54

26 Traffic Data Collection and AnalysisC apter 4Counting Characteristics

Roads Department

t s es rab e t at t e ro ng-out software creates a e n a e system

that permits unlimited numbers of new parameters to be added on.

he data from the manual counting stations can be retained in the manual

forms or be transforme to t e ata e system w ere computers are

available for further processing and proper storage. A special computer

program w ave to be eve ope for t s purpose. W ere off t e s e ve

software are available such software should be used instead of developing

new programs.

rom the automatic traffic counters data should be transferred from the

count ng equ pment to a PC or to a ata mo u e n t e form eve ope

to suit the format available within the institution/organisation. With the

proper system n p ace, t en t e ata can automat ca y be converte to

the file system on a PC as specified by the relevant authority.

ata from t e automat c traf c count ng system s ou be comp e sepa-rately for traffic flow directions.

t is recommended that the same file system both for manual and auto-

mat c count ng ata s ou be use to ensure atta nment of t e requ re

comparison of information or results.

he data should be stored as files in the specified file system that has a

e name n or er to perm t easy recogn t on of ata. For secur ty reasons

cop es of es s ou be store as z ppe es, but cr t ca y, ana yse ata

should be stored with a password in a secured place in the computer system.

urt er, t s conven ent to eep t e bac up of t ose es preferab y n

D-R M.

4.1.3 Routine Counts

ese are norma or sc e u e counts con ucte at regu ar nterva s for a

set purpose, such as monitoring of traffic flow patterns for determination

of stor ca tren or eterm nat on of app cab e factors e.g. growt or

conversion factors).

4.1.4 A -Hoc or nsche ule ountsAd-hoc counts may be conducted a ong some roads, wh ch carry ow traf c

ntensities. However, over the years as it becomes evident that traffic on

these roads has increased, scheduled traffic counting stations may be

estab s e .

4.1.5 Origin Destination Surveys

n an Origin and Destination (O-D) survey, interviews of vehicle users by

enumerators are con ucte . T e approac to suc O-D survey epen s

on the scope and quality of informations required. Some of the methods

ava ab e nc u e roa s e nterv ews, postcar , tag an sc, ve c e

registration number etc.

r or to commencement of O-D survey, t e enumerators s a e r efe

about t e tas an f necessary, a fam ar sat on course s ou be un er-

taken before commencement of such surveys. The survey shall be carrie

Institutions engaged in traffic data collection

for development and maintenance of the road

network should used readily available systems

r programs.


27/54

27Traffic Data Collection and Analysis apter 4Counting Characteristics

Roads Department

out at ea ocat ons av ng goo v s b ty w t out av ng prom nent

orizontal and vertical curves. Further, selected spots shall preferably

ave a equate safety str p to par t e ve c es away from t e traf c ane.

Depending on the requirement the enumerators shall be instructed to collect

nformation on the following:

me of Interv ew.

ype an we g t of commo ty be ng carr e .

ype of vehicles.

Number of passengers n t e ve c e.

r g n of t e ourney.

mmediate destination.

na est nat on.

tart time of journey. Expected end time of journey.

requency of tr p.

ther relevant informations.

he origin and destination points are the starting and final destination

zones of a tr p, w c may be n fferent zones as we as n t e same

zone.


28/54

28 Traffic Data Collection and AnalysisChapter 5Resources Required for Collectionof Traffic Data

Roads Department

RESOURCES REQUIRED FOR COLLECTION

OF TRAFFIC DATA

5.1 Generalssessment of ava ab e resources pr or to commencement of any act v-

ty is critical to any assignment at hand. For traffic data collection, it is

mportant t at proper assessment of t e extent or scope of t e env sage

counting (quality level of data required) is undertaken. This is aimed at

ensur ng t at t e p anne an organ se exerc se s ac eve at opt ma

cost and with the expected accuracy. The exact number of persons an

equ pment to un erta e a spec c traf c count ng ass gnment s epen-

ent, among ot ers, on: T e ocat on of t e stat on.

The quality of data to be collected.

T e eve of traf c ow.

T e nature of t e roa sect on an traf c ow c aracter st cs w t n

which the station falls.

Traf c compos t on.

5.2 Staff Composition and Qualification

More often staffing for conducting traffic counts is not deemed to becr t ca to t e qua ty of ata be ng co ecte , as t e exerc se s regar e

ust as count ng of ve cu ar traf c an groups as t ey pass a spec c

point on the road. While taking cognisance of this concept, the selection

cr ter a conta ne ere n s not prescr pt ve as t on y recommen e t at

n the case of Botswana, field enumerators appointed for traffic surveys

s ou ave at east a m n mum of Genera Cert cate of Secon ary

Education (GCSE) to enable them to be continuously trained in this field.

enumerators upon emp oyment s ou un ergo a tra n ng programme

before being assigned to work alone in the field. This training is aime

at fam ar s ng new emp oyees w t ve c e c ass cat on/con gurat on

processes, use of different automatic counting equipment as stated early

where available , refer to Guideline No. 4,Axle Load Surveys any otheract v ty assoc ate w t traf c ata co ect on an ana ys s.

W e t e ma or ty of t e enumerators may be base n t e e co ect-

ng raw data, some should be stationed at office to transform the raw

ata nto ana ys s forms, e t er for subsequent storage or transfer nto t e

computer system to further analysis.

e of ce staff s ou ave a wor ng exper ence as e enumerators to

ensure full understanding of the raw data they are expected to transform

nto the analysis forms. They should as well have at least a GCSE Cer-

t cate. T e Of ce Enumerators s ou un ergo a t oroug Computer


29/54

29Traffic Data Collection and Analysis Chapter 5Resources Required for

Collection of Traffic Data

Roads Department

software tra n ng programme, suc as ata process ng, sprea s eets,

appropriate traffic data analysis system(s) available and any other relate

systems for t e ex gency of t e serv ce.

ue to the changes in the types of the vehicle fleet using the road networ

an t e vary ng qua ty an use of traf c ata, traf c enumerators s ou

ave specialised courses arranged at set periods or intervals. The courses

s ou nc u e, among ot ers, t e mportance of traf c ata co ect on

an ts use, emp oyment regu at ons, rst a programmes, an any ot er

programme deemed to be relevant for the achievement of the service or

ass gnment. However, spec a se courses s ou be ta or ma e for of ce

enumerators n comput ng an t e management of store ata.

5.3 Equipment Requirement

5.3.1 Automatic Traffic Counting Equipmentlthough a detailed discussion of the various types of automatic traffic

counter equ pment an systems was ea t w t n Chapter 2.0t e ma n

aim of this part is to sight typical equipment required for traffic data

co ect on. In essence, typ ca automat c traf c ata co ect on equ p-

ment consists of a etector to etect vehicles an a counter to recor the

nformation, some of which are shown below.

Figure 5.3.1: Typical layout of traffic loops on the road surface.

Figure 5.3.2: Typical equipment for traffic data collection.

A security fence for traffic counting equipment.

Roadsidecabinet

Piezo electric detectorfor weigh in motion

2end set of loops to

measure speed andlength (for classification)

Counting equipment fixed to a traffic signpost.

A roadside box or cabinet for storage of atraffic counter.


30/54

30 Traffic Data Collection and AnalysisChapter 6Traffic and Site Safety

Roads Department

Road Dep rt e

6 TRAFFIC AND SITESAFETY

6.1 Generalraf c safety ur ng t e con uct of traf c surveys s man atory an s

the responsibility of the institution or body undertaking the surveys. The

oad Traffic Act Cap 69:01 places a statutory responsibility in ensur-

ng t at appropr ate safety measures are n p ace before a survey can be

conducted on a road.

6.2 Site Safetyhe location of the counting sites should be chosen with full consider-

at on to traf c safety ot for nsta at on, ma ntenance an use of t esite. Whenever manual traffic surveys are in progress, proper signals

s ou be n p ace for safety of enumerators. T e s te s ou be nspecte

for safe use by supervisor who should also ensure that no sign is remove

from s te unt t e survey s comp ete .

6.3 Site Markingsoth the automat c and manua count ng s tes shou d have an unamb guous y

dentification number. The counting site number is linked to the exist-

ng Roa Reference System. It s recommen e t at t e counter storage

boxes be marked with an identity number plate.


31/54

31Traffic Data Collection and Analysis Chapter 7Traffic Counting Procedures

Roads Department

Ro ds Dep r me t

7 TRAFFIC COUNTINGPROCEDURES

7.1 GeneralThe result of traffic counting is sub ect to sampling error and observational

uncerta nty. Samp ng error n traf c count ng s error emanat ng from co -

ected traf c data wh e observat ona error re ates to veh c e c ass cat on

by veh c e types resu t ng n some veh c es be ng wrong y c ass ed. In

t s context, ve c e c ass cat on cannot e e ne w t out am gu ty an

therefore is a subject of enumerators interpretation of the passing traffic

stream. To minimise the error, statistical methods are more preferable to use

or analysis to smooth out sampling and observational errors.

utomat c counters ma n y use t e stance between ax es to c ass fyve c es. In s tuat ons w ere ve c es of fferent ma e ave s m ar ax espac ng, t e automat c counter cannot ascerta n t at t ese are two fferent

veh c es. The resu t ng scenar o s that e ther the system re ects a veh c es

t cannot c ass fy or m sc ass es t em, t us resu t ng n c ass cat onerror. Where there is a significant proportion of unclassified vehicles thelevel of error is deemed to be very high and the results should not be useor any economic decision making purpose.

typ ca s tuat on s t at of Toyota 4 x 4 H ux an Toyota H Ace comb ,ave t e same ax e spac ng. It s t erefore a v sab e t at traf c ata from

an automat c counter s cross c ec e by ta ng a manua count to enab e

easy checking and comparison of the error margin.

7.2 Factors Affecting Vehicle Countinghere are many factors that affect traffic counting and the most common

nc u es:

Weather conditions.

Purpose of the traffic counting.

Met o of traf c count ng.

Location of the counting sites.

Traf c ow eve .

Road type.

Traf c compos t on.

Weather Conditions

raf c count ng ur ng g an ow weat er con t ons ave ser ous ef-ect on t e enumerators t roug an nterrupte count ng process resu t ngn severe ata errors.

Purpose of Traffic Counting

Understanding the purpose for which traffic is being counted enables collec-

tion of appropriate quality data, the number of vehicle classes and determi-

nation of the duration of the count. For major projects, good quality data is

required for a pre-determined number of vehicle classes at specific intervals

over at east ve work ng days for 12 to 16 hours per day, exc ud ng pub c

ho days or spec a occas ons, such as ndependence day. Any other exerc se

that is not of major significance should require less counting duration.

A Toyota 4x4 Hilux has similar axle spacing asa Toyota Hi-Ace combi/minibus.

A Toyota Hi-Ace combi/minibus.

Location of traffic counting sites is a function

of the end use of data being collected.


32/54

32 Traffic Data Collection and AnalysisChapter 7Traffic Counting Procedures

Roads Department

Method of Traffic Counting

he method of traffic counting has a bearing on the quality of data obtained.

anua count ng of veh c es pass ng a po nt at spec c nterva s 15 m nutes,

0 minutes, 45 minutes, 1 hour, etc.) is one method of traffic counting,

whereby the tota veh cu ar traf c s requ re . If from the past traf c counts

t is required that a proportion of heavy vehicles is to be determined as a

percentage of t e tota traf c, t en suc a percentage s app e to t e

counted volume to get the effective traffic flow. On this basis, the metho

of traf c count ng s ou be ec e before an an t at traf c count ng

forms and training should be conducted.

Location of Counting Sites

Counting in the open country and in build-up areas is different in nature

an n execut on. Traf c count ng n an open country rura env ronment

as a high potential to yield the much-needed concentration by enumera-

tors, w e t at n t e bu -up areas s prone to sturbances. T e type o

disturbances envisaged include movement by enumerators between thecount ng s tes an t e bu t-up env ronment, suc as s ops, t us resu t-

ng n g error marg ns n ata so co ecte . It s, t erefore, cr t ca t at

supervision and close monitoring are undertaken at the respective sites

fa ng w t n t e bu t-up env ronment for t e urat on of t e counts.

W ere traf c count ng s to be con ucte for an ntersect on, t e number

of enumerators required is mainly a function of the type of intersection,

t e compos t on of traf c ow, t e number of anes an t e ant c pate

traffic volume. It is, therefore, essential that a proper and detailed site

nspect on s con ucte pr or to start of any traf c survey.

Traffic Flow LevelTraffic flow level influences the capability of enumerators to carry out traffic

counts manua y on h gh vo ume roads. Where the enumerator s expected to

count more than one traffic lane on a busy road, observational error is com-

monly encountered. On the other hand, whereas efficiency and capability o

traf c enumerat on vary from enumerator to enumerator, for a traf c eve

of less than 1000 vehicles per hour in one direction manual counting by one

enumerator can g ve sat sfactory data w th n the perm ss b e to erances. For

traffic level in excess of 1000 vehicles per hour in one direction, additional

enumerators are requ re or automat c counters s ou e use .

Figure 7.2.1: Different average daily traffic flow levels.

A traffic counting station in a rural area.

Normal continuous traffic flow condition.

0

2000

4000

6000

8000

10000

12000

14000

DailyTrafficflow

Mond

ay

Tuesda

y

Wedne

sday

Thursday

Frida

y

Saturda

y

Sund

ay

Days of the week

Congested or forced traffic flow condition.


33/54

33Traffic Data Collection and Analysis Chapter 7Traffic Counting Procedures

Roads Department

Road Type

e type of roa as an effect on traf c count ng as t comb nes bot t e

level of traffic flow and the number of lanes for the specific road. Traffic

ow norma y fo ows t ree fferent mo es an eve s, name y:

Congested or forced low-speed continuous flow;

Normal continuous flow and;

W e space g -spee ow.

rom t e ower eve to t e upper eve of traf c ow t ere rema ns an

ncremental desire for collection of better quality data commensurate with

ncrease n t e samp e s ze. T s s because w t ower vo ume of traf c

the potential for variance in the traffic flow reduces, while with higher

traf c vo ume potent a for ow var at on over s ort nterva s ncreases,

suc as t a ow or p atoon ow.

W e ta ng cogn sance of t e above traf c ow eve s, traf c count ngor single carriageways can be handled with manual counting with much

re uce r s of error eve w e ua carr ageways wou requ re auto-

matic traffic counters owing to the volume of traffic, the speed and the

number of anes.

Traffic Composition

raf c compos t on n uences t e met o o ogy an approac w c

should be applied for traffic counting at respective counting sites. This

ctates t e number of ve c e c asses an t e qua ty of ata requ re .

t is, therefore, critical that the traffic composition is known prior to the

commencement of t e survey n or er to enforce t e measures requ re .

7.3 Counting Accuracy and QualityAssurance

uality of sampled data largely depends on, among others, the methoollowed in the establishment, use and maintenance of the count stations.or automatic counts equipment, close co-operation with the manufac-

turer s necessary n or er to ac eve max mum bene ts. T e operat onproce ures s ou be c ear y wr tten an eta e for easy un erstan ng

y ot t e c ent an t e supp er.

Common for all methods of traffic counting, classification chart (baseon vehicle composition) is required to record the traffic flow by vehicle

types. This is because traffic flow is a dynamic process and can fluctuateover time. Accuracy in traffic counting and establishment of these fluctua-tions in the traffic flow is of critical importance, as this influence deriva-t on of pro ecte traf c. W ere t ere s ev ence of ncrease traf c ow,t s requ re t at t e ve c e groups exper enc ng g growt s ou be

ca cu ate as a percentage of t e tota traf c counte n t e stream. T sena es appropr ate eterm nat on of t e mpact of t e most n uent avehicle type(s) on the road being considered and how it influences eco-nomic analyses and decision-making.

Where different classes of traffic characteristics shows a differentialgrowth in the stream, average growth rates for each traffic class shall be

ca cu ate as t e average growt rate for t e tota traf c counte .

Widespread or platoon/tidal traffic flowondition.

Mixed traffic flow condition.


34/54

34 Traffic Data Collection and AnalysisChapter 7Traffic Counting Procedures

Roads Department

s t ese ca cu at ons can exper ence errors, t t en emp as ses t e nee

to undertake proper training before start of traffic counting exercise. In

a t on, ong cont nuous counts can ave cumu ate uman error ue to

at gue and ack o concentrat on by enumerators, wh ch w then a ect the

resu ts of t e ca cu at ons. On t e ot er an , t e s ort urat on counts

can result in overestimation of the assessment ue to insufficient ata.

ese s ou be guar e aga nst n or er to ensure cons stency n ata

collection and reduced potential for human error.

n t s context, t e user of t e traf c ata s ou now t e purpose for

which data is being collected and its level of accuracy. Attempts should,

t erefore, be ma e n ensur ng t at t ere s a equate p ann ng, tra n ng o

enumerators and a clear understanding of the exercise to be performed.

n order to compensate for possible road dependant factors, calibration

fac t es are des rab e. Some automat c count ng equ pment have automat c

ca brat on, w e ot ers can on y a ust t e sens t v ty or to erance. Someave the arrangement to adjust the loop length using different vehicles

w t nown engt s. T e ca brat on rout nes ave to be con ucte ac-

cording to the instruction given by the manufacturer of the equipment.

he accuracy of traffic counting depends on its duration and variation in

t e traf c ow. T e fo ow ng examp es of accuracy eve s for automat c

counting are used in Norway, which can be adopted to Botswana condi-

t ons after proper stat st ca ca rat on for t e 90t an 95t percent e

con ence eve s.

- 6 wee s count ng: 4.9 % an 5.8 %

- 4 weeks counting: 7.4 % and 8.7 %- 2 wee s count ng: 9.7 % an 11.5

- 1 week counting: 16.2 % and 19.3

s the seasonal variation of the traffic flow is very low in Botswana, the

engt of t e count ng per o s can be re uce as eeme appropr ate

depending on the quality and end use of the data.

7.4 Duration of Countinguration of traffic counts is dependent on the type and quality of data

requ re . Depen ng on t e en use of traf c ata be ng co ecte , counts

at established permanent stations are ideally conducted over 12, 16 and4 ours cont nuous y for at east seven consecut ve ays per stat on. T e

selection of counts duration will depend on whether the amount of data

co ecte w pro uce re ab e resu ts. T s ec s on w epen on t e

characteristics of the traffic flow and the type data required for a particu-

ar location or project. Typical duration are given in Table 2.3.1 as shown

ear er n t s gu e ne.

Much consideration needs to be taken into account when planning a traffic

survey an t e re at ve mportance of t ese w vary w t t e purpo

Botswana_Guideline 9 - Traffic Data Collection and Analysis (2004)

Documents

t oug t s gu e ne s

w t t e d str

ng roa s

enab e t em

preva ng eve s of traf

c t zencontractors

facilitateby t e pub

counc s