1. Institut Superieur De Technologies. Burkina Faso.

*             Correspondence: dariusniyomugabo11@gmail.com

Keywords: analysis, modelling, sustainable, urban, public transport, performance, Kigali city

1. INTRODUCTION

In the space of just a few decades, urban areas across the world, in both developed and developing countries, have become increasingly automobile-dominated and less sustainable [1]. Rwanda is one of developing countries, Kigali city other cities in developing countries have experienced a rapid growth in transport-related challenges, including accident, congestion, pollution, public transport decline, environmental degradation, climate change, depletion of energy, visual intrusion, and lack of accessibility for the urban poor [2]. Achieving sustainability of the transportation system is a critical component of Sustainable development [3]. According to [4], there is almost no place left on the planet where humans have not had a major impact on the plant and animal life. In the more densely populated areas, our built environment has thoughtful effects on such things as rain runoff and ground water levels. And natural habitat is progressively cut up into smaller pieces, while industrial agricultural and forest lead to enormous mono-cultures [5]. Past research was estimated that urban areas shelter less than 3% of Earth’s landmass, but in terms of ecological footprint [6,7]. Transportation is connected to all aspects of human life. According to [8], our natural environment, economic prosperity, and social wellbeing all be contingent on transportation systems that are harmless, clean, efficient, and equitable. Though, current predictions suggest that transportation growth is unsustainable. It looms our environmental, economic, and social future. Altering and managing trends in transportation presents a significant challenge and will require the cooperation of all stakeholders at local, regional, national and international levels. According to [9], the provision of adequate and appropriate public transport services is one of the most important components for wellbeing of growing and expanding. World Bank reported that an innovative approach is needed to meet urban mobility challenges especially in African cities. Maximum of these cities are exploited by a high population growth, inadequate transport and extreme poverty [10]. The rapid population growth and urbanization, coupled with increasing economic activities and opportunities in the cities, result in rapidly growing travel demand, both for private as well as public transport [11].  To provide accommodations of this rapid growth in the demand for private transport requires very extensive road capacity, which would induce more greenhouse gas emissions. Otherwise, public transport is adequate for mass mobility; it makes better use of urban space, reduces the reliance on more polluting modes of transport, and is likely to be a reasonable means of transport for most residents in these cities. Delivery of passable and appropriate public transport services is one of the most important components for comfort of growing and expanding urban areas [12]. Experience has shown that, public transport has a great significance in reducing traffic congestion, offering alternative means of travel, and contributing greatly to the quality of urban life [13]. Public transport is a public service, and should provide service levels that comply with public demand [14]. With a growing population and rapid urbanization, public transport systems need to be updated as well [15]. An interval between growing public transport demand and service capacity results in an increase of travel cost, congestion, and unreliable service, thereby creating economic loss and environmental degradation [16]. Therefore, it is of substantial value that approaches for monitoring, assessing and modelling public transport system performance are developed, in order to ensure a provision of better services. The study area of this research is focused on Kigali. Kigali city is the national capital and the most important business Centre in Rwanda. It is characterized high altitude, in a tropical climate with a hilly countryside sprawling across ridges and wet valleys in between. The rapid urbanization in the Kigali city had resulted in unplanned settlements, urban extension, and increasing urban poverty. Subsequently, [17] reported that more than 80% of the population of Kigali lives in informal neighborhoods. Sustainable public transport is a strategy designed to meet the mobility needs of people and businesses in cities and their close areas for a better quality of life. It builds on existing planning practices and takes due account of the principles of integration, participation and evaluation. Sustainable transportation systems contribute positively to the environmental, social and economic sustainability of the communities they serve. Transport systems exist to provide social and economic connections, and people are quick to seize the opportunities offered by increased mobility, with poor households benefiting greatly from low-carbon transportation options. The benefits of increased mobility need to be weighed against the environmental, social and economic costs of transport systems. The basic purpose of sustainable public transport has been to provide mobility to people without access to private cars and to meet present and future need. Nowadays, public transport is adopted for many purposes, such as providing mass mobility, managing traffic jams, and creating development opportunities [18]. In other side, Sustainable transportation provides access to all groups of people in the city in a manner that is within the environmental carrying capacity of the city and is inexpensive to both the providers and the users of the system. Public transport plays a vital role in creating competitive economies, livable, and inclusive communities within the city to (State of Western Australia). Additionally, many authors have also expressed that public transportation is one of the potential ways of mitigating air pollution, easing traffic congestion, reducing energy consumption, and improving mass mobility [1,7]. The ability to improve sustainability of public transport services relies on the ability to measure the public transport system performance. The planning and operation of public transport service are closely tied to the ability to measure the spatial and temporal accessibility of the public transport systems [12]. Therefore, this increases the need for assessment of public transport network routing and service capacity available to meet the demand. According to [14], Planning and assessment of a sustainable urban public transport system, especially in terms of routes to be operated, and components of services capacity to meet transit demand is a composite task, and it requires a considerable analysis. In this case Geographical Information Systems (GIS) is greatly used as transport models and technologies such as in urban transport planning [8]. GIS have the ability to integrate maps and spatial analytical methods, which make it a powerful tool for transport planning. Not only that but also transport information system (TMS) designed to provide end to end logistic management process ticket booking and vending and travel. GIS are proving to be valuable transportation management and modelling plat-forms [19]. However, this study is trusted on information from GIS based approaches to develop Kigali public transport model. Moreover, objectively verifiable indicators are needed to assess the sustainability of public transport system of Kigali. This study analyzed and modeled the Sustainable Urban Public Transport Performance in Rwanda: Case study Kigali City Region.

2. MATERIALS AND METHODS

The research area includes Kigali city as a capital city of Rwanda and a city with high population compared to other city in country, population in Kigali increasingly day to day the total population reached 1,132,686 people with a population according to 2012 Rwandan census.

 

Figure 01: Population growth in Kigali Source: world population growth prospect (2019)

Figure 02: Location of Kigali city

This study is based on the literature review, case study observation and using quantitative assessment. It reviews the theoretical aspects of research work of sustainability factors and performance in urban transportation. The data collections are regained from the Kigali city, Government institutions, transportation agencies and other resources. Most map data were collected in Excel format. Intended for analysis in Arc GIS, these data were transformed into ESRI shape file. Numerous data are found in shape format in open data website and other resources. The following data were obtained from the official government and several sources are open data. Digital maps of Kigali Boundary are in the form of GIS files, road centerlines, road network polygons, road names, and road types, collected from website address: gis.bpbd.kigalicity.go.id. Demographic data, such as population number, population growth, total car number, total car ownership, total school, income per district, level of education, employed and unemployed, and transportation statistics are collected from the website: kigalicity.bps.go.id and Rwanda’s National Institute of Statistics (NISR). Transportation data, such as traffic congestion and traffic accident, are collected from police traffic corps office. Traffic air pollution and traffic noise pollution data are collected from Rwanda Environmental Management Agency. Documents such as transportation master plan are from transportation ministry website and other sources. This study uses several methods of GIS approach to model the performance of Sustainable Urban Transport in Kigali city with five indicators and provided by ArcGIS version 10.5.1 (ESRI, Redlands, US). Therefore, this study has selected five basic indicators of sustainable urban transport data, namely traffic congestion, traffic accident, traffic air pollution, traffic noise pollution, and land consumption for transport infrastructure (see figure4). Recognizing the relationship between indicators is very important for transport planners and urban planners. GIS helps describe and classify any type of features, such as points, line sand polygons based on values of attributes’ data [19]. A diversity weighting approach that evaluates the SUT model for each district within the study area was used based on the previous research by Ramani et al. (2009). Each indicator is classified into five weights: High = 1 (lowest problems), Medium–High = 2, Medium = 3, Medium–Low = 4, and Low (highest problems) = 5 using natural breaks type of GIS classification tool.

Table 01: Data collection years and sources

Table 02: Five basic indicators of index formulas

Basic Indicator of SUT	Equations
Traffic Congestions Indicator (TCI)	TCI= Total congestion point/area
Traffic Accident Indicator (TACI)	TACI=Total accident number/ area
Traffic Air Pollution Indicator (TAPI)	TAPI= the level of api1/area
Traffic Noise Pollution Indicator (TNPI)	TNPI=the level of NPI2/area
Transport Infrastructure in Land Consumption (TILCI)	TILCI= Total road area/district area

 

Figure 03: SUT: Sustainable Urban Transport; LU: Land Use; GIS: Geographic

In GIS, spatial statistics is used to help assess patterns, trends, and relationships for better understanding behavior of geographic phenomena, pinpoint causes of specific geographic patterns, make decisions with higher level of confidence and summarize the distribution in a single number. Spatial statistics tools in ArcGIS have several functions [20]. According to [21], Moran’s Index test statistics is given by the following equation:

Another important GIS tool in analysis based on spatial statistics is spatial regression analysis [14]. OLS as a global regression model has one equation for all features which is calibrated using data from all features, while the type of relationship is fixed. Whereas, GWR is a local form of linear regression used to model spatially varying relationships with each feature has one equation calibrated using data from nearby features; the type of relationships can vary across the study area. This study models five basic indicators of Sustainable Urban Transport Performance using spatial statistics as shown in Figure 04.

 

Figure 04: Modelling Relationship of Sustainable Urban Transport Performance Index (SUTPI) indicators

Therefore, this study proposes the regression models for Sustainable Urban Transport Performance Index (SUTPI) as follows.

SUTPI = β0 + β1(TCI) + β2(TAccI) + β3(TAPI) + β4(TNPI) + β5(TILCI) + ε

Where SUTPI as dependent variable (Y): variable to model or predict, five basic indicators as explanatory variables (X): variable that influence or help explain the dependent variable.

 

Figure 05: Modelling Relationship of SUTPI indicators

3. RESULTS AND DISCUSSION

Figure 06: Traffic Congestion at a given Roundabout in Kigali

Figure 07: victims in accident by categories of persons in 2017

Figure 08: deceased persons by categories of vehicle in 2017

Figure 09: Fatal accident and serious injury accident happened in 2017

Accident occurrences are high in Kigali City compared with countryside highways. The major cause of accidents is that there are many motorcycle taxis in Kigali City. The traffic accidents by motorcycles account for 58% of total traffic accidents, which is a problem in the urban area. Death rates are higher in the countryside than in Kigali. The reason being that there are few crackdowns by police and thus, vehicles move at high speed and in addition, sidewalks are not installed in the countryside. The death of pedestrians by accident accounts for 40% of total death by accidents, and improvement is necessary. The number of traffic accidents is decreasing because    speed governors are being installed.

 

Figure 10: victims in accident by categories of persons in 2018

Figure 11: deceased persons by categories of vehicle in 2018

Figure 12:  Fatal accident and serious injury accident happened in 2018

 

Table 02: Approach to Estimate emissions

Table 03: Emission inventory of Road Transport Sector of Kigali

Table 04: Vehicle categories contributing maximum towards emissions of different pollutants

 

Figure 13: Land Use Distributions

Table 05: Jurisdictional distribution of Kigali Districts

No	District names	Total area(km2)	Total population
1	KICUKIRO	167	318,564
2	GASABO	430.3	529,561
3	NYARUGENGE	154	281,561

 

Table 06: Summary of Four Exclusive Public Transport Zones

Table 07: Kigali population distribution by age

Age Cohort	Population	Percentage	Cumulative percentage
0 – 14	314959	36.5	36.5
15 – 29	324461	37.6	74.1
30 – 44	148663	17.2	91.3
45 – 59	52761	6.1	97.4
60 – 74	16293	1.9	99.3
75+	5738	0.7	100

 

Figure 14: Kigali 2008 population densities

Table 08: Rwanda road network

RWANDA ROADS	Length (km)	Road density (km/km2)
Classified Road
Total paved national roads	1075	0.04
Total unpaved national roads	1785	0.07
Total unpaved District roads	1838	0.07
Total classified roads	4,698	0.18
KIGALI CITY ROADS
Total paved road in Kigali	864	0.21
Total unpaved roads in Kigali	511	1.18
Total Kigali City roads	1375	1.39
Rural feeder roads
Total unpaved feeder roads	8285	0.31
Grand total of roads in Rd	14,000	0.53

Table 09: The major road network is shown in the map below

Road class	Road Name	Connected major cities	Remark
National highway connected to Kigali	RN1	BUTARE	West direction
	RN3	BYUMBA	North direction
		RWAMAGANA	East direction
	RN4	RUHENGERI	Northwest direction
	RN5	BUGESERA	South direction

 

Figure 15: Major Road Network in Kigali

Table 10: Classification of Roads in Kigali City in 2018

Earth Road	Road Asphalt Pavement	Concrete Pavement Road
2,400 km	428 km	23 km

 

Figure 16: classification of road in Kigali

Table 11: Cumulative Number of Vehicles Registered by Categories

 

Figure 17: Growth Rate

The organized survey reported that availability of footpaths was as low as 7% of the total road length. It was also found that only 6% of the total population actually uses the footpath and remaining 94% avoids it. This resulted due to two main reasons, as poorly maintained footpaths, and the lack of continuity. Hence, absence of infrastructure restricts users from walking even the shorter distances of up to 1 to 2 kms. Moreover, it is tongue-in-cheek to find that when it comes to modes, walking is widely opted. But the infrastructure available is not even half the total road length. Therefore, to promote walking among all the categories of occupants it becomes imperative to first provide the adequate infrastructure facilities in the city. Parking spaces are installed in commercial facilities such as shopping malls and hotel, and public facilities such as government offices. There are public parking lots between road and buildings in the city. When the driver parks Kigali Veterans Cooperatives Society (KVCS) staff collects the parking fee. Many vehicles are parked on the street in central urban area because the parking lot is insufficient.

 

Figure 18: Parking Space

The parking fee is set according to Presidential Order N25/01. As shown in Table 4.8 below, parking fees are paid per category of vehicles either per hour, per day, or per month.

Table 12: fees charged on parking

 

Figure 19: Main Bus Terminal

The above figure shows the main bus terminal in Kigali. Those ARE 9 terminals as Nyabugogo, Kacyiru, Kimironko, Remera, Kabuga, Nyanza, CBD. Some traffic management systems such as traffic signals and roundabouts can be found in the urban areas. The locations of traffic signals and roundabouts are shown in Figure 4.7 below. Currently, most of the intersections in the city have not been signalized yet.

 

Figure 20: Locations of Signalized Intersection and Roundabout

There are six signalized intersections in the city (Giporoso, Gishushu, Kabean, Sopetrad, Peage, and Gakingiro) which operate on fixed timing. Each is a standalone signal with no communication between each signal. In the nighttime from 10:00 p.m. to 6:00 a.m., the system is switched into a flashing signal. The Rwanda National Police (RNP) usually control the traffic flow using hand signal at the congested intersections during peak hours. All of the signal devices such as signal controller, pole, light, and cable are made in China.

Figure 21: Signal Controller and Cable

There are 12 roundabouts in Kigali City. The roundabout adjacent to the convention center is partially closed due to security concerns. Traffic capacity is insufficient during peak hours in some of the roundabouts and it becomes a cause of traffic congestion.

 

Figure 22: Traffic Congestion at the Roundabout

 

 

Table 12: Traffic Accident-Related Articles

Closed-circuit television (CCTV), Vehicle Enforcement System, and Traffic Control Center are installed and operated by RNP. CCTV was installed in Kigali City two years ago for monitoring traffic accidents and crime prevention. There are 190 CCTVs installed in 52 locations. There are two kinds of camera, namely, fixed camera and pan tilt zoom (PTZ) camera. The Vehicle Enforcement System is a system that consists of speed detectors and cameras, and detects excessive speeding and also signals ignoring vehicles. This system also includes Track Management System which monitors violating vehicles using information on license plates captured by cameras installed at several places. Thirty-six cameras are planned to be installed, and five are currently in use. The remaining 31 cameras will be installed in 2018. In case of speeding violation, the police can identify the driver from the information on the license plate and send SMS to the mobile phone. Offender who receives SMS message shall pay a penalty charge through the mobile money system or to the bank account of RNP. At the Traffic Control Center stationed at the RNP, the police officers use the cameras shown above to monitor the speed violations and traffic accidents on a 24-hour schedule. At least 21 screens are installed at the Traffic Control Center. There are two police officers to monitor traffic violations, and 13 police officers to monitor crime. In addition, 12 officers are stationed in the call center to provide guidance to the field sites. It is noted that the center of RNP is not responsible for traffic management such as control of the traffic signal system but responsible for the enforcement of crime including traffic-related violations. The transport planning process is usually carried in a number of sequential stages, in what has been called four-stage classic transport model. The approach starts with considering a network and zoning system, called a traffic analysis zone (TAZ), and the collection of data. These data are used to estimate a model of the total number of trips generated by or attracted to each zone: Trip generation. The next stage is concerned with the estimation of the number of trips per unit time which will be made under certain circumstances between each pair of zones in an area to which the process is being applied: Trip distribution. The following stage is usually the Modelling of the choice of mode to be used for making the trip: Modal split. The last stage is to provide an estimate of the amount of traffic which will use each part of a transport network under certain conditions: Trip assignment [21,22]. This four-stage transport model is based on gravity model concept; the estimated pattern of trips depends on the trip magnitude between each pair of zones, and is inversely proportional to the costs of travel between the various pairs of zones.

 

Figure 23: The classic four-stage transport model

Table 13: Future Road Development Plan

Table 14: Future Development Plan

The undertaken research is solely interested on the estimation of potential trips for public transport in Kigali. Based on a 24 hours survey carried out by Japan Engineering Consultants in 2004, the vehicle origin-destination (OD) in Kigali was 58,700 trips consisted of 46.6% by private cars, 29.4% by buses and 24% by trucks (Japan Engineering Consultants, 2004). According to this survey, the average numbers of passengers carried by private car, bus, and truck, was respectively 2.6, 15.4 and 3. These figures provide the generated trips for public transport in 2004. Due to the lack of socio-economic data, we assumed that the production of trips per cell, which is considered as a TAZ, being proportional to the population of that cell. According to [23], a trip rate per capita may be used for forecasting public transport demand in a typical urban area. In line with this respect, we assumed that the trip rate per capita in 2004 will apply in 2011, and hence total volume changes only in response to population growth. Likewise, the estimation of attracted trips per TAZ was affected by the lack of socio-economic data, and the fact that Kigali has no land use map ever produced. We assumed that the trip attraction per cell is proportional to the population density. This was supported by the fact that 60% of Kigali population work in informal sector, and the fact that Kigali is comprised by unplanned and squalid settlements resulted from the rapid urbanization, with more than 80% of the population living in informal neighborhoods [24]. Moreover, the population density is an overall measure of intensity of activities, including residential, employment, and all other activities, assuming that they are generally closely correlated [25,26]. Finally, the trip attraction is modified by a matching factor to ensure balance between the total number of trips produced and attracted in the city. To rationalize the trip assignment to the network, the trips generated at each TAZ were overlaid with small hexagons of 400-meter edge length, in order to arrive at trip productions and attractions at spatially disaggregated level. The distribution trip was estimated from average growth rate method using the base year model OD. The average growth rate method is a prediction of the future distribution trips by calculating the average growth of trips in both the generation zone and attraction zone.

Figure 24: Modal Split Ratio except NMT and Freight Traffic

Figure 25: BRT Station’s Territories

In addition to the BRT, the current bus service is planned to be continued. The number of urban bus users was estimated considering the BRT development and distributed to each traffic zone.

 

Figure 26: Trip Growth Ratio by Mode of Transport

Mode share helps in understanding the trip load shared by a particular mode in a study area. It is the ratio of trips made by a particular mode of transport to the total trips made in the study area. According to the primary survey, almost half of the total trips that are being made by the city are done by non-motorized means of transport with 49% and private vehicle and public transport has approx. equal share of 26% and 25% respectively. Although this doesn’t provide a comprehensive share of trips by different modes of transport, therefore a detailed share of each mode. During the survey process, respondents were referred to know what changes they would want in the public transport services to increase its popularity and usage. The replies evidently highlighted the priorities between Slum and non-Slum residents. Firstly, the Slum respondents prefer the same cost of travel, but ask for higher frequency and reduction in the travel time by ½ or ¼. Surprisingly, comfort was not valued much during the travel, which is concluded based on their preference of still opting for the non-AC buses.  Secondly, the non-Slum respondents provided an important insight about the kind of services required. Majorly, the two opted preferences were, first, higher fare for the AC buses with higher frequency and reduced travel time; and second, to keep the same fare for the low frequency buses i.e. only during the peak hours with reduction up to ¼ of the total travel time and need of AC buses. Hence, preference given to AC buses by the non-slum respondents’ shows importance of comfort during their travel. While analyzing the average travel time by individual mode and mode preferences, it is clear that higher time taken by the public transport is a perceived notion by the residents who prefer private vehicle. This because both private and public modes of transport take similar amount of time to travel the same distance. Transit accessibility delineates the ease to access bus stop and IPT stop from residence. A stop within 0.5 kms radius of a residence is the benchmark for transit accessibility. On the basis of this, the slum and non-slum surveyed households are checked for accessibility to transit. The transit accessibility for each category of household is examined for each zone. The two Figures 5.19 and 5.20 below show the accessibility to bus and IPT stop in different district of Kigali by the non-slum and slum category of the residents in the city respectively. The purpose of this analysis is to understand the availability of transit in proximity with the residence and thus establishing a relation between the level of usage and travel expenditure incurred by the slum and non-slum household. Furthermore, the comparison between bus and IPT stop is made to understand the popularity of a particular mode of transit. There are four public transport zone in Kigali the covered area of that zone is described as follow.

ZONE1 include the area such as Remera, Kanombe, Kabeza, Nyarugunga, Rusororo (Kabuga), Masaka, and Ndera Sectors. ZONE2 include the area of Niboye, Kicukiro (Sonatubes, Centre), Gahanga, Gatenga, Gikondo and Kigarama. ZONE3 Comprise area of Kimironko, Kinyinya (Kagugu & Dutchwelle), Gisozi, Kacyiru, New Gakinjiro, Batsinda, Kibagabaga, Kimihurura, Nyarutarama ZONE4 covers the area of Kimisagara, Nyakabanda, Nyamirambo, Mageragere, Kigali, Gatsata, Karuruma, Jabana, and Nyacyonga.

 

Figure 27: Transit Accessibility of Non-Slum population

Figure 28: Transit Accessibility of Slum population

The model considered the trips uniquely generated for public transport mode. Hence there is no modal split stage applied in our modelling process. This section only explains the steps followed to assign the trips to public transport network.  In earlier stages while constructing the bus route network as shown in Figure 13, we constructed route segments by joining intersection points of bus routes in order to later allow overlaps of different routes. This resulted in 78 segments in total. However, since we had to assign the trips to network based on node in Flow map software, we split the bus routes by bus stops in ArcGIS. This resulted in 228 small segments, named route splits. These route splits were converted into Flow map file to represent the public transport network. The idea is simply to allow a more accurate assignment of the flow over the network. In this respect, a node in Flow map is likely to represent the real bus stop.

Table 15: Total length and trip times

 

Figure 29: Result of the Traffic Demand Forecast A) 2025, B) 2030 and C) 2040

As the dependent variable, Sustainable Urban Transport Performance Index (SUTPI) had mean value approximately 2.98 (almost 3, meaning in the middle performance level) and standard deviation of 1.42. For the explanatory variable i.e., the mean of traffic congestion indicator was about 2.67 point per district with standard deviation of 1.32. The mean value of traffic accident indicator was 3.14 and standard deviation of 1.39. For traffic air pollution indicator had mean value was 2.76 and standard deviation of 1.34. Traffic noise pollution indicator had mean value of 2.88 with standard deviation of 1.26. Land consumption indicator had mean value of 3.21 and standard deviation of 1.34 (see Table 16).

Table16: Descriptive Statistic (based on classification scale from 1 to 5).

 

Figure 30: Variable distributions and relationships. (a) SUTPI and TCI; (b) SUTPI and TAccI; (c) SUTPI and TAPI; (d) SUTPI and TNPI; (e) SUTPI and TILCI

Table 17: The coefficients of SUTPI Model.

Figure 31: (a) Standard Residuals of OLS SUTPI (b) Non-clustered Residual

Table18: Summary of Variable Significance.

 

Table 19: GRW Analysis Result.

 

Figure 32: Non-Clustered Residual

GRW is an essential step in modeling methodology that the simulation data and the real data should be compared in order to validate the model and calculate the relative error and mean square deviation of the indicators. In the model validation stage, there are two important aspects, namely the suitability of the behavior pattern between real data and the simulation results and about the close relationship between the values of real data and the simulation results.  This model formation is the consequence of the displaying procedure of the causal impact relationship of different affected segments whereas; the precision level and accuracy of the model are reflected by the closeness between the real data values and simulations results. The examined variables on this study consist of five indicators: traffic congestion, traffic accident, traffic air pollution, traffic noise pollution, and land consumption for transport infrastructure. GRW analysis resulted the predictive model of SUTPI, three basic indicators were significant predictors, namely traffic congestion, traffic accident and traffic air pollution. Two basic indicators were not significant predictors, namely traffic noise pollution and land consumption. There was no multi-collinearity problem of the predictive model since the Variance Inflation Factor (VIF) ranged under 7.5. Furthermore, the adjusted R-square of the model was 0.845 indicating that 84.5% variability in SUTPI value could be explained by these variables.

Table 20: SUTPI Predicted Performance in Kigali city based on GRW analysis

Performance	SUTPI		Zone Name
	Total zone	Area (Km2)
High	7	10	Remera, Kimironko, Kimisagara, Nyamirambo, Kicukiro (Sonatube,centre), Nyabugogo, Gakinjiro
Medium–High	8	12	Kanombe, Kabeza,Niboye, Gatenga, Gikondo, Nyacyonga, Batsinda,Kibagabaga
Medium	8	12.35	Gahanga, Kigarama, Niboye, Kimihurura, Nyarutarama, Gisozi, Kacyiru,Gahanga
Medium–Low	4	5	Nyarugunga, Ndera, Kinyinya,nyakabanda
Low	6	8.7	Rusororo, Masaka, Kabeza, Karumuna, Mageragere, jabana,

 

Figure 33: GRW Model Validation between SUTPI observed and SUTPI predicted

The studies that are pure in exploring the measurement of SUT performance are still rare in recent times, especially those which use a quantitative approach and spatial analysis. For instance, there are current studies about sustainable urban transport index which have been developed by UNESCAP (2018) that introduce the measurements of SUT in four Asian cities, but the study results are still generally qualitative in describing the indicators presented by the spider diagram. In addition, the study explored by Doust and Parolin using two indicators, namely accessibility and greenhouse gas emission in Sydney city, used the metric methodologies as the core the study for displaying the analysis result whereas this study uses a simple approach, more specific and spatially to assist the transport and urban planners in identifying transportation issues for making strategic plans of SUT. Further studies can create SUT performance scenarios by first conducting the calibration process and clearly defining the sustainable urban transport performance index (SUTPI) by including measurement units to clarify how to read and interpret the results. The City of Kigali should initiate a Car Free Day event. One or more road corridor may be closed and made free of vehicular traffic. This event should be introduced with an aim of “Encouraging people to walk, jog and cycle as part of Active Transport and health lifestyle promotion. The move is aimed at encouraging mass sports and exercise while introducing the much-awaited green transport and green city initiatives as per the City of Kigali’s goals.

Figure 34: Mass sport and exercise in Kigali

The following elements are critical in the improvement of public transport service in Kigali city.

• More than 22 new routes may be created following 41 initial routes and the existing ones extended to reach residential areas in a bid to reduce walking distances to the nearest bus stops.
• Each transport zone should be dedicated operator who is held accountable in case of poor service delivery.
• A number of bigger buses should be purchased replacing the former smaller minibuses which serve the feeder routes.
• An introduction of internet in buses and E-Ticketing was done based on the existing fiber internet propagation

Improvement of the existing main Corridors from single to dual carriage ways in order to initiate dedicated bus lanes for PT priority in peak hours and upgrade of the cross sections by adding green space, footpath and cycle path where applicable

 

Figure 35: Sona tube Roundabout

 

 

Table 21: Elements of intelligent transport system

Efficient Road Network	Vehicle use Management	Better Public Transport
ü  Traffic management ü  Re-routing guidance ü  Ramp metering ü  Variable speed controls ü  Incident detection and management	ü  Access control ü  Road user charging ü  Congestion charging	ü  Journey planning ü  Real-time passenger information system ü  Buss traffic priority

 

Figure 36: Intelligent Transport System

The green transport city will increase sustainability of urban public transport in Kigali by considering 70% public transit modal share and 10% green trips (non-motorized trips).

Figure 37: Non-Motorized Transit, Green network, cycling, Dedicated Space for Pedestrians and Bicycles

4. CONCLUSION

The measurement of sustainable urban transport performance is difficult through the evaluation of its indicators, but the Sustainable Urban Transport Performance Model (SUTPM) technique is able to prove to be a reliable method. This method is better than previous techniques which mostly counts one indicator of SUT and mostly qualitatively. The SUTPM method is designed to have more capability in measuring the performance of SUTs spatially and simply. The indicators used are by using basic indicator in local and regional scope. This model is to visualize the effect of the indicator on the performance of the SUT and its influence respectively. Therefore, this model can be used to measure and evaluate the development of urban transport in order to be sustainable. Spatial patterns of travel demand and service distribution in each district are different. Each district has a performance value from the SUT to be measurable for its development. For the global study result indicates that the performance of SUT in Kigali in medium level. The level of traffic congestion, air pollution and noise pollution continue to show increasing trends, caused by population growth, and the rapid growth of the use of private vehicles. Although, on the other side, accident rate indicates trends that continue to decrease and land use for road infrastructure in general is not excessive. This research has investigated the sustainable urban transport conditions in the city of Kigali with the basic indicators. It has identified the degree of sustainability in the urban transport system. Comprehensive and robust results can be obtained by analyzing those indicators: traffic congestion, accident rate, air pollution, noise pollution, and land use for transportation infrastructure. However, it is necessary to repair the completeness of the data and the latest version and using network analysis in modeling approach to get more specific in result and precise prediction.

REFERENCES

1. Agency New Zealand Transport. (2010). Public transport network planning. a guide to best practice in NZ cities research report, 396.
2. Armstrong-Wright, A. &. (1987). Bus services: Reducing costs,Raising standards. urban transport series, The World Bank.
3. Barrella, E. M. (2012). STRATEGIC PLANNING FOR A SUSTAINABLE TRANSPORTATION SYSTEM: A SWOT-BASED FRAMEWORK FOR ASSESSMENT AND IMPLEMENTATION GUIDANCE FOR TRANSPORTATION AGENCIES . Georogia: Georgia Institute of Technology august 2012.
4. Beimborn, E. H. (1999). Land use: transportation interaction. In an overview, Land use and economic development in statewide transportation planing. Milwaukee: Federal highway administration.
5. Ben-Elia, E. S. (2003). Transportation Impact Statment(TIS): A new tool for transportation and land use planning. Environment and Planning, 2177-2190.
6. Bennet, L., & Vale, F. (2014). Modeling Spatial Relationships Using Regression Analysis; ESRI: Redlands,CA. USA.
7. Best, A., Giljum, S., Simmons, C., Blobel, D., Lewis, K., Hammer, M., Cavalieri, S., Lutter, S. & Maguire, C. . (2008). Potential of the ecological footprint for monitoring environmental impacts from natural resource use: Analysis of the potential of the ecological footprint and related assessment tools for use in the eu’s thematic strategy on the sustainable use of natura. Report to the european commission, DG Environment.
8. Camagni, R. G. (2002). Urban mobility and urban form: the social and environmental costs of different patterns of urban expansion. Ecological Economics, 199-216.
9. Davison, L. J. (2006). Bus quality partnerships, modal shift and traffic decongestion. . Journal of transportation Geography,, 177-194.
10. Diaz Olvera, L. P. (2008). Household transport expenditure in sub-saharan african cities: measurement and analysis. Journal of transportation geography, , 1-13.
11. dorina and Dominic. (2015). urban transport sustainability. sustainability.
12. Geerlings, H. K. (2006). Development of a methodology for benchmarking public transport organization: a practical tool based on an industry sound methodology.
13. Hall, R. P. (2002). Introducing the Concept of Sustainable Transportation to the U.S. DOT through the Reauthorization of TEA-21 . Massachusetts: Massachusetts Institute of Technology,June 2002 .
14. Hensher, &. B. (2000). Handbook of transportation modelling(Vol.1). Amsterdam: Eliseviere Science.
15. Horner, M. W. (2004). Spatial representation and scale impacts in transit services assessment.Environment and Planning B. Planning and Design, 785-797.
16. Jonsson, R. D. (2008). Analysing Sustainable Urban Transport and Land-Use ;Modelling tools and appraisal frameworks. SE-100 44 Stockholm, Sweden: Royal Institute of Technology .
17. Lewin, K. L. (1939). Patterns of Aggressive Behavior in Experimentally Created Social Climates. Journal of Social Psychology, 10, 271–301.
18. (2008). Rwanda Transport Sector Policy. 28.
19. Murray, A. T. (1998). Public Transportation Access. Transportation Research Part D: Transport and Environment, 3(5), 319-328.
20. NIYONSENGA, D. (2012). ASSESSING PUBLIC TRANSPORT SUPPLY FOR KIGALI, RWANDA . Enschede, The Netherlands: University of Twente .
21. Puji Adiatna Nadi and AbdulKader Murad . (2019). Modelling Sustainable Urban Transport Performance in the Jakarta city Region: A GIS Approach. Sustainability.
22. Rees, W. and Wackernagel, M. (1996). Urban ecological footprints: Why cities cannot be sustainable and why they are a key to sustainability. Environmental Impact Assessment Review, 16(4-6), 223-248.
23. stead, D. p. (2015). In the space of just a few decades, urban areas across the world, in both developed and developing countries, have become increasingly automobile-dominated and less sustainable. In developing countries in particular, cities have experienced a rapid growth. sustaibability.
24. Transport, A. N. (2010). public transport planing. a guide to the best practice in ZN cities research report, 396.
25. Vuchic, V. R. (2005). Urban Transit: Operations, Planning, and Economics. John Wiley & Sons, Inc. .
26. Zuidgeest, M. H. (2011). Introductionry notes in transport planning and travel demand modelling. lecture book. Urban planning and Management ITC.Enschede.