Zacharia Lukorito Chepkania Page 7 of 109
Chapter 1
Introduction
1.1 Overview
Kenya continues to experience high traffic volume flow on its roads. Coupled with this is
the rise in the number of road crashes reported annually (Assum, 1998). Many agencies
such as the traffic police, the ministry of roads and public works, the health ministry and
the municipal councils need these data to identify problems, plan on the allocation of the
funds to improve road construction and hospital locations and support development and
evaluation of highway and vehicle safety.
For any meaningful analysis to be achieved, data needs to be collected over some period
of time to establish the trend of occurrences. In Kenya this has not been maximized since
most of these data is still paper-based and stored in large cabinets making retrieval and
analysis a great problem. Often these data is never used at all, making the road safety
countermeasures a hurdle. Lack of analysis of these data has made Kenya rank among
countries reporting high road crash rates in the world (Assum, 1998).
Having realized the importance of crash data and its analysis as means of mitigating road
crashes and reducing fatalities, the government of Kenya has acknowledged that crash
data system needs to be put in place to provide these vital, yet often neglected
information. It is in this light that the road safety unit under the ministry of roads and
public works embarked on this project, to have the data required by various agencies
readily available and in a manner that can be retrieved easily. It has been envisioned that
crash data will be utilized to determine cost effective improvements of the roads, reduce
crash frequencies, fatalities and injuries and improve response time in the event of a road
crash.
Potential benefits underlie crash data analysis. From the analysis, additional causes of
road crashes are identified. Only then can better, informed decisions be made to remedy
existing or potential hazardous locations. It also enables engineers and planners to design
Zacharia Lukorito Chepkania Page 8 of 109
and plan safer roadways and neighbourhoods by providing a clearer picture of
contributing factors in certain crashes or crash types. This is strengthened by Khayesi
(1997) who acknowledges that using the analysis from the crash data the police and heath
agencies are guided on the allocation of emergency response (e.g., police, ambulance and
hospitals) in specific locations.
While considering the crash system, it is imperative to look at the geographical relations
to correlate the crashes and physical locations. Geographic Information Systems (GIS)
provide excellent tools to analyse location-specific crash data. Multiple layers can be
viewed and analysed at once. GIS’s enable development of a methodology to consider
non-roadway variables.
1.2 Crash Data
Traditionally, the police collect and store the road vehicle crashes mainly for prosecution
purposes and insurance claims. The police use the P41 form (shown in Figures 1.1 and
1.2) to record all the road crash parameters. Some of the information recorded is fairly
static and predetermined, such as the type of vehicles involved hence these are sometimes
coded. The other is largely written long hand. The reporting by the police is the only
comprehensive source of data on road crashes in Kenya and is used by many agencies in
the country (Odera et al., 2003).
Once the police in various administrative regions collect the data, it is submitted to the
traffic headquarters for entry onto an MS Excel sheet, as the front-end application
analysis. The review of the traffic statistics for the year 2001 to 2003, contained
information on crash severity (Fatal, serious or slight), categories of road users affected
type of vehicles involved and causes of crashes. This information is partly given in
appendix 4. One deficiency identified is that it takes long for the reports to be compiled
and made available. Secondly, the information provided is not exact as there is no proper
linkage between the hospitals and the police. For instance, if death occurs within one
month of the crash, then the crash is considered fatal. Unfortunately, this is never
Zacharia Lukorito Chepkania Page 9 of 109
reported to the police for updating. Therefore there is likelihood that numbers of fatalities
are being under-reported, a view that is strongly shared by Khan et al. (2004).
There has been no formal system for location referencing and it has been acknowledged
as a fundamental deficiency in the crash records. Crash location is still being described in
words in general terms with often a road name and a landmark being used.
There is no formal exchange of crash data between police and other agencies except for
some high level reporting to the Ministry of Transport who have a standing requirement
to produce some general quarter yearly statistics. Transfer and exchange of information
by the police is done over radio and in most cases no exchange of information is done.
1.3 A Brief Description of the System
In this project, I have developed a Geographic Information System (GIS)-based
integrated road crash data system. This is a computer-based road crash data analysis
system utilizing the basic relational database system and the ArcView GIS. Its purpose is
to enable Kenya collect and analyse road crash data in a systematic manner with easy-to-
use software. The program is specifically designed for use on a personal computer, which
enables users to work interactively with ready accessible data (something which was not
always possible with the paper-based system which is still being used today). The system
has two distinct sections. The first deals with the input of crash data from police road
traffic accident report forms based on MS access, while the second deals with the
location analysis of the crash data in ArcView GIS.
The system allows crashes with common causes to be identified. Crash, casualty and
vehicle tabulations can be produced using a number of queries, which can be designed in
the system. In addition, users can make their own tabulations.
With the development of the crash data system more accurate recording, storage, location
referencing, updating and exchange of crash information between different agencies will
be achieved accurately and timely.
Zacharia Lukorito Chepkania Page 10 of 109
Figure 1.1 — Sample form P41 (page 1)
Zacharia Lukorito Chepkania Page 11 of 109
Figure 1.2 — Sample form P41 (page 2)
Zacharia Lukorito Chepkania Page 12 of 109
1.4 Project Aims and Objectives
The main project aim is to design and implement a prototype crash data system based on
MS access software and the ArcView GIS that allows users to analyse crash data and use
this information to develop countermeasures towards improving road safety in Kenya.
The system should be easy to use, with a clear layout and should have adequate
functionality, for example by carrying out crash causes analysis.
As well as requiring an interactive system, the road safety unit would also like to use the
system to produce reports for the ministry of transport. In the future, the functionality of
the reporting system could be enhanced by allowing information such as individual crash
reports being reported securely to ensure confidentiality.
As I have no prior knowledge of ArcView GIS, learning ArcView was another objective.
The project terms of reference are included in Appendix 1 and outline the project
objectives in more detail.
1.4 The Structure of the Thesis
The structure of the thesis is presented as follows:
Chapter 1 has been the introduction chapter. The chapter starts with the description of the
general problem of the crashes and the increase of the traffic flow in Kenya. It then
focuses on the need for the analysis of the crash data, briefly analysing the deficiencies in
the current crash recording, storage, analysis and reporting system. The chapter then
briefly looks at what the crash data system is, thereby formulating the problem at hand
and setting the project aims and objectives.
Chapter 2 reviews the basics of database systems, the available data models and sets the
stage for the use of the relational database system as the preferred platform for the crash
data system attribute data management.
Chapter 3 reviews the GIS applications crash analysis with the benefits they offer over
and above the paper-based system. The review indicates that the GIS offer improvements
Zacharia Lukorito Chepkania Page 13 of 109
in data management and analysis of crash data. It provides the different agencies with the
information needed for improvements of the road conditions and networks.
Chapter 4 describes the system development, mainly focussing on the database design
and integration with the ArcView GIS. The research concepts reviewed in chapters 2 and
3 are applied here. The data modelling techniques using the entity relationship diagrams,
as well as the relational database rules are used in the design of the database system using
MS access. The integration of spatial and non-spatial data is presented here with the
linkage being established between the MS access database and the ArcView GIS.
Chapter 5 describes the integration of the theoretical research with the practical
development of the system. Consideration is given to the extent of which the research has
been applied in the design and development of the practical project objective.
Chapter 6 evaluates the project overall success. Consideration is given to the user
requirements and involvement, project management and success of the project as per
deliverables achieved. An attempt has been made to show how each of the project
objectives has been achieved.
Chapter 7 summarises the project and gives the overall conclusion and lesson learned and
as well as presents further research areas.
1.6 Summary
This chapter has introduced the general overview of the system and presented the
problems experienced by the use of the paper-based system. It has been found that paper-
based systems take large volumes of storage space, they are not easy to manipulate,
making it difficult to extract the required data. This presents to both the user of the
system and the agencies that require such information with inaccurate and incomplete
information. In such cases analysis is incomplete and decisions arrived at based on such
analysis are haphazard. In the long run the problems are never solved.
Ricevi informazioni sui nostri servizi, sulle offerte e non perdere news e consigli su università e lavoro.
