Driving behavior considerably affects the transportation system, especially lane-changing behavior occasionally cause conflicts between drivers and induce shock waves that propagate backward. A freeway merge section is one of locations observed a freeway bottleneck, generating freeway traffic congestion. The emerging technologies, such as autonomous vehicles (AVs) and vehicle connectivity, are expected to bring about improvement in mobility, safety, and environment. Hence the objective of this study is to enhance freeway merge section operations based on the advanced technologies. To achieve the objective, this study modeled the non-cooperative merging behavior, and then proposed the cooperative applications in consideration of a connected and automated vehicles (CAVs) environment. As a tactical process, decision-making for lane-changing behaviors is complicated as the closest following vehicle in the target lane also behaves concerning to the lane change (reaction to the lane-changing intention), i.e., there is apparent interaction between drivers. To model this decision-making properly, this study used the game theoretical approach which is the study of the ways in which interacting choices of players. The game models were developed to enhance the microscopic simulation model representing human driver's realistic lane-changing maneuvers. The stage game structure was designed and payoff functions corresponding to the action strategy sets were formulated using driver's critical decision variables. Furthermore, the repeated game concept which takes previous game results into account was introduced with the assumption that drivers want to maintain initial decision in competition if there is no significant change of situations. The validation results using empirical data provided that the developed stage game has a prediction accuracy of approximately 86%, and the superior performance of the repeated game was verified by an agent-based simulation model, especially in a competitive scenario. Specifically, it helps a simulation model to not fluctuate in decision-making. Based on the validated non-cooperative game model, in addition, this study proposed the cooperative maneuver planning avoiding the non-cooperative maneuvers with prediction of the other vehicle's desired action. If a competitive action is anticipated, in other words, a CAV changes its action to be cooperative without selfish driving. Simulation results showed that the proposed cooperative maneuver planning can improve traffic flow at a freeway merge section. Lastly, the optimal lane selection (OLS) algorithm was also proposed to assist lane selection in consideration of real-time downstream traffic data transferred via a long-range wireless communication. Simulation case study on I-66 highway proved that the proposed OLS can improve the system-wide freeway traffic flow and lane allocation. Overall, the present work addressed developing the game model for merging maneuvers in a traditional transportation system and suggesting use of efficient algorithms in a CAV environment. These findings will contribute to enhance performance of the microscopic simulator and prepare the new era of future transportation system. / Doctor of Philosophy / Driving behaviors considerably affect the traffic flow; especially a lane change occasionally forces rear vehicles in a target lane to decrease speed or stop, hence it is considered as one of primary sources causing traffic congestion. U.S. Department of Transportation (DOT) announced that freeway bottleneck including merge section contributes to freeway traffic congestion more than 40 percent while traffic incidents count for only 25 percent of freeway congestion. This study, therefore, selected a freeway merge section, where mandatory lane changes are required, as a target area for the study. The emerging technologies, such as autonomous vehicles (AVs) and vehicle connectivity, are expected to bring about improvement in mobility, safety, and environment. Based upon these backgrounds, the objective of this study was determined to enhance freeway merge section operations based on the advanced technologies. To achieve the objective, first this study focused on understanding driving behaviors of human drivers. Decision-making for lane-changing behaviors is complicated as the closest following vehicle in the target lane also behaves concerning to the lane change (reaction to the lane-changing intention), i.e., there is apparent interaction between drivers. For example, the vehicle sometimes interferes the merging vehicle's lane-changing by decreasing a gap. To model the decision-making properly, this study modeled the non-cooperative merging behaviors using a game theoretical approach which mathematically explains the interaction (e.g., cooperation or conflict) between intelligent decision-makers. It was modeled for two vehicles, i.e., the merging vehicle in acceleration lane and a following vehicle in freeway rightmost lane, with possible actions of each vehicle. This model includes how each vehicle chooses an action in consideration of rewards. The developed model showed prediction accuracy of approximately 86% against empirical data collected at a merge section on US 101 highway. This study additionally evaluated the proposed model's rational decision-making performance in various merging situations using an agent-based simulation model. These evaluation results indicate that the developed model can depict merging maneuvers based on practical decision-making. Since most existing lane-changing models were developed from the standpoint of the lane-changing vehicle only, this study anticipates that a lane-changing model including practical decision-making process can be used to precisely analyze traffic flow in microscopic traffic simulation. Additionally, an AV should behave as a human-driven vehicle in order to coexist in traditional transportation system, and can predict surrounding vehicle's movement. The developed model in this study can be a part of AV's driving strategy based on perception of human behaviors. In a future transportation environment, vehicle connectivity enables to identify the surrounding vehicles and transfer the data between vehicles. Also, autonomous driving behaviors can be programmed to reduce competition by predicting behaviors of surrounding human-driven vehicles. This study proposed the cooperative maneuver planning which future connected and automated vehicles (CAVs) avoid choosing the non-cooperative actions based on the game model. If a competitive action is anticipated, in other words, a CAV changes its action to be cooperative without selfish driving. Simulation results showed that the proposed cooperative maneuver planning can improve traffic flow at a freeway merge section. Lastly, the optimal lane selection (OLS) algorithm was also proposed to provide a driver the more efficient lane information in consideration of real-time downstream traffic data transferred via a long-range wireless communication. Simulation case study on I-66 highway proved that the proposed OLS can improve the system-wide freeway traffic flow and lane allocation. Overall, the present work addressed developing the game model for merging maneuvers in a traditional transportation system and suggesting use of efficient algorithms in a CAV environment. These findings will contribute to enhance performance of the microscopic simulator and prepare the new era of future transportation system.

The purpose of this research is to develop methodologies for applying traffic flow theories to various ITS categories through the utilization of evolving surveillance technologies. This integration of theory, measurement and application has been overlooked since the advent of ITS because of the number of disciplines involved. In this context, the following illustrative methodologies are selected, developed and presented in this study: - a methodology for automatic measurement of major spatial traffic variables for the present and the future implementation of various ITS functional areas, in general; and - a methodology for real-time link and incident specific freeway diversion in conjunction with freeway incident management, in particular. The first methodology includes the development of a dynamic flow model based on stochastic queuing theory and the principle of conservation of vehicles. An inductive modeling approach adapted here utilizes geometric interpretations of cumulative arrival-departure diagrams which have been drawn directly from surveillance data. The advantages of this model are real-time applicability and transportability as well as ease of use. Analysis results show that the estimates are in qualitative and quantitative agreement with the empirical data measured at 30-second intervals. The analytical expression for link travel times satisfies traffic dynamics where the new form of the equation of conservation of vehicles has been derived. This methodology has potential applicable to automatic traffic control and automatic incident detection. The methodology is then applied to freeway diversion in real-time in conjunction with freeway incident management. The proposed new form of the equation of conservation of vehicles is applied to detect recurring or non-recurring congestion analytically. The principle of conservation of vehicles is applied to develop the concept of progression and retrogression of incident domain, which turns out to be compatible with traditional shock wave traffic mechanism during incidents. The link and incident specific diversion methodology is achieved by using a delay diagram and volume-travel time curves, which can be plotted per link per incident. The use of such graphic aids makes problem solving much easier and clearer. The dynamic traffic flow model developed here can also be applied to estimate travel times during incidents as a function of time. The development of a computer program for freeway diversion concludes this research. / Ph. D.

Thesis (MA)--Stellenbosch University, 2015. / ENGLISH ABSTRACT: Mega projects have fascinated human beings for as long as history can remember. The urge to build something bigger and better than has ever been done before has always been a driving force behind the human race’s relentless pursuit of technological advancement. It is in this vein that mega projects have evolved over time, as methods of construction improved, so did the scale on which people could attempt new projects. The Channel Tunnel between England and France, the Hoover Dam in America and the Millau Viaduct in the French countryside are some of the biggest examples of infrastructure projects in the world and these are all unequivocally, mega projects. Costing at least $ 250 million and incorporating major technological challenges, mega projects continue to inspire and motivate artists and engineers alike. This thesis seeks to expand people’s understanding of the analysis of these mega projects. Mega project analysis is a field that has struggled to differentiate itself ordinary project analysis. The Gauteng Freeway Improvement Project (GFIP) and the associated e-tolling mega project will be analysed in this thesis. The framework for analysis will be provided by the work of Flyvbjerg, Bruzelius and Rothengatter (2003), who seek to analyse mega project success or failure based on three key indicators of economic sustainability, environmental concerns and the effect of public support. This thesis will use the indicators of economic sustainability and the effect of public support to determine whether the Gauteng Freeway Improvement Project (GFIP) can be viewed as a failed mega project or not. After an in-depth study of the data and material available, this descriptive and explanatory study shows that the GFIP and associated e-tolling mega project is indeed a failed mega project. This is because it has failed the analysis in both categories of economic sustainability and the effect of public support. / AFRIKAANSE OPSOMMING: Megaprojekte het al eeue lank die mens gefasineer. Die neiging van die mens om iets groter en beter te maak as wat al ooit gemaak was het nog altyd baie dryfkrag verskaf vir die mens se strewe na tegnologiese verbetering. Dit is met hierdie idee in gedagte dat megaprojekte met tyd verander het, soos wat boumetodes verbeter het, so ook het die grootte van projekte verander wat mense kon aanpak. Die Channel Tonnel tussen Engeland en Frankryk, die Hoover Dam in Amerika en die Millau Brug in die Franse platteland is voorbeelde van die grootste infrastruktuur projekte ter wêreld en hierdie is al drie, sonder enige twyfel, megaprojekte. Teen ‘n koste van ten minste $ 250 miljoen en met grootskaalse tegnologiese uitdaging, hou megaprojekte aan om vir beide kunstenaars en ingenieurs te motiveer en uit te daag. Hierdie tesis poog om mense se kennis van die analise van megaprojekte te verbreed. Megaprojek analise is ‘n veld wat al jare lank sukkel om verskille te bewerkstellig tussen homself en gewone projek analise. Die “Gauteng Freeway Improvement Project” (GFIP) en die verwante e-toll megaprojek sal geanaliseer word in hierdie tesis. Die raamwerk vir analise sal deur Flyvbjerg, Bruzelius en Rothengatter (2003) verskaf word, waar die outeurs poog om megaprojekte se sukses of mislukking te bepaal gebaseer op die sleutel aanwysers van ekonomiese volhoubaarheid, omgewingskwessies en die effek van openbare ondersteuning. Hierdie tesis sal gebruik maak van die ekonomiese volhoubaarheid en openbare ondersteuning aanwysers om te bepaal of die GFIP beskou kan word as a mislukte megaprojek of nie. Na ‘n in-diepte study van die data en materiaal beskikbaar, sal hierdie beskrywende en verduidelikende studie wys dat die GFIP en verwante e-toll megaprojek inderdaad ‘n mislukte megaprojek is, as gevolg daarvan dat die GFIP megaprojek analise aandui dat die megaprojek misluk het in beide die ekonomiese volhoubaarheid en publieke ondersteunings aanwysers.

Thesis (MDF)--Stellenbosch University, 2011. / The aim of this research report is to analyse the feasibility of Open Road Tolling (ORT) and its development in South Africa through the Gauteng Freeway Improvement Project (GFIP). ORT represents the next generation of Electronic Toll Collection (ETC) and this research report will assess to what extent the GFIP scheme is in line with other comparable tolling schemes; and is the institutional environment amenable to ORT. This will allow one to gauge the feasibility of the scheme and its potential for acceptability and success. The research report outlines the number of risks that come with an ORT scheme and these include amongst others collection risk, enforcement, technology, privacy and public acceptance. The success of the GFIP will largely be determined by how well these risks are mitigated and how the benefits can be marketed to the users. The literature review illustrates that whether road pricing schemes have failed to move forward, have been implemented, are currently under development, or still in the planning stage as a concept there are several consistent lessons and critical success factors one should apply when structuring a scheme. In the discussions with stakeholders, the following conclusions with regards to the feasibility of ORT and its development in South Africa were as follows: The factors which need to be addressed include political risk, effective marketing of the scheme to the public, obtaining political will and support, building trust between the scheme developer and the user, managing perceptions and acknowledgement of the fact that the scheme will need to prove itself over time. Inadequate demonstration of equity for the user along with poor communication would compromise public acceptance and the success of the scheme. Incorporating interoperability yields benefits in terms in terms of network externalities, the ability to use a single transponder for multiple tolling plazas and points, along with the potential for alternative uses for the transponder. ORT as a viable solution for the GFIP is feasible from a technical point in that it's the only way in which one can collect tolls from a high volume network and not cause disruptions in the flow of traffic. However, there are a number of persistent residual risks that SANRAL cannot entirely mitigate and some fall under the realm of political risk. While SANRAL has applied best practice principles in structuring the GFIP with the aim of providing value for money for the user and as far as possible tackling the issue of affordability, there are certain realities, such as the recent global financial crisis, the infrastructure backlog of the country, users paying for roads which were free and challenges with overall service delivery which place a strain on the legitimacy of the GFIP ORT scheme.

Simuladores de tráfego são ferramentas importantes para planejamento e operação do tráfego, pois permitem avaliar cenários de maneira segura e com baixo custo. Porém, a configuração original desses programas é ajustada para replicar as condições do tráfego observado no país de origem do simulador. Em rodovias, o modelo de desempenho de caminhões tem grande influência no comportamento da corrente de tráfego. No Vissim, todavia, esse modelo simula veículos com melhor desempenho, se comparados aos caminhões brasileiros. Apresenta-se neste trabalho a calibração das funções de aceleração para caminhões do Vissim, utilizando a diferença entre perfis de velocidade simulados e reais como medidas de ajuste. Os perfis de velocidade foram criados a partir de dados de GPS para caminhões que trafegaram ao longo de uma rodovia de pista dupla em relevo ondulado, sob baixo fluxo de tráfego. O ruído na velocidade presente nesses dados foi atenuado por meio de um filtro gaussiano. Determinou-se então o perfil longitudinal da rodovia através dos pontos coletados pelo GPS, de modo a ser possível comparar as velocidades reais e simuladas em pontos preestabelecidos ao longo do trecho. A calibração foi automatizada por meio de um algoritmo genético. Todas as replicações da calibração resultaram em configurações para as funções de aceleração capazes de gerar perfis de velocidade simulados mais próximos dos reais se comparadas à configuração original do simulador. Por fim, um estudo de caso foi usado para demonstrar a necessidade de recalibração do simulador, através da avaliação do impacto causado pela calibração do simulador nos equivalentes veiculares obtidos com dados simulados. / Traffic simulators are important tools for traffic planning and operation, as they can be used to perform safe, low-cost scenario evaluation. However, these software\'s default configuration is adjusted to replicate the traffic conditions found in the simulators\' country of origin. The truck performance model heavily impacts traffic behavior in simulations of freeway segments. The default values for Vissim\'s performance model are adjusted for trucks with lower mass/power ratios than typical Brazillian trucks. The objective of the research reported here was to recalibrate Vissim\'s truck acceleration functions using the difference between real and simulated truck speed profiles as goodness-of-fit measures. The speed profiles were created using GPS data obtained from trucks travelling over a four-lane freeway situated on rolling terrain, under low traffic flow. Speed noise was removed from GPS data using a Gaussian filter. The longitudinal profile of the freeway was determined using GPS data, to allow for the comparison of observed and simulated speeds at fixed points along the segment. The calibration procedure was automated and based on a genetic algorithm. Every calibration run provided a configuration for the acceleration functions that reduced the difference between simulated and real speed profiles, if compared to the simulator\'s default configuration. The effects of the truck performance model recalibration were demonstrated by means of a case study focusing on the estimation of passenger-car equivalents for Brazilian trucks travelling on freeways.

Simuladores de tráfego são ferramentas importantes para planejamento e operação do tráfego, pois permitem avaliar cenários de maneira segura e com baixo custo. Porém, a configuração original desses programas é ajustada para replicar as condições do tráfego observado no país de origem do simulador. Em rodovias, o modelo de desempenho de caminhões tem grande influência no comportamento da corrente de tráfego. No Vissim, todavia, esse modelo simula veículos com melhor desempenho, se comparados aos caminhões brasileiros. Apresenta-se neste trabalho a calibração das funções de aceleração para caminhões do Vissim, utilizando a diferença entre perfis de velocidade simulados e reais como medidas de ajuste. Os perfis de velocidade foram criados a partir de dados de GPS para caminhões que trafegaram ao longo de uma rodovia de pista dupla em relevo ondulado, sob baixo fluxo de tráfego. O ruído na velocidade presente nesses dados foi atenuado por meio de um filtro gaussiano. Determinou-se então o perfil longitudinal da rodovia através dos pontos coletados pelo GPS, de modo a ser possível comparar as velocidades reais e simuladas em pontos preestabelecidos ao longo do trecho. A calibração foi automatizada por meio de um algoritmo genético. Todas as replicações da calibração resultaram em configurações para as funções de aceleração capazes de gerar perfis de velocidade simulados mais próximos dos reais se comparadas à configuração original do simulador. Por fim, um estudo de caso foi usado para demonstrar a necessidade de recalibração do simulador, através da avaliação do impacto causado pela calibração do simulador nos equivalentes veiculares obtidos com dados simulados. / Traffic simulators are important tools for traffic planning and operation, as they can be used to perform safe, low-cost scenario evaluation. However, these software\'s default configuration is adjusted to replicate the traffic conditions found in the simulators\' country of origin. The truck performance model heavily impacts traffic behavior in simulations of freeway segments. The default values for Vissim\'s performance model are adjusted for trucks with lower mass/power ratios than typical Brazillian trucks. The objective of the research reported here was to recalibrate Vissim\'s truck acceleration functions using the difference between real and simulated truck speed profiles as goodness-of-fit measures. The speed profiles were created using GPS data obtained from trucks travelling over a four-lane freeway situated on rolling terrain, under low traffic flow. Speed noise was removed from GPS data using a Gaussian filter. The longitudinal profile of the freeway was determined using GPS data, to allow for the comparison of observed and simulated speeds at fixed points along the segment. The calibration procedure was automated and based on a genetic algorithm. Every calibration run provided a configuration for the acceleration functions that reduced the difference between simulated and real speed profiles, if compared to the simulator\'s default configuration. The effects of the truck performance model recalibration were demonstrated by means of a case study focusing on the estimation of passenger-car equivalents for Brazilian trucks travelling on freeways.

Rainfall affects the performance of traffic operations and endangers safety. A common and conventional method (rain gauges) for rainfall measurements mostly provide precipitation records in hourly and 15-minute intervals. However, reliability, continuity, and wide area coverage pose challenges with this data collection method. There is also a greater likelihood for data misrepresentation in areas where short duration rainfall is predominant, i.e., reported values may not reflect the actual equivalent rainfall intensity during subintervals over the entire reporting period. With recent weather and climate patterns increasing in severity, there is a need for a more effective and reliable way of measuring rainfall data used for traffic analyses. This study deployed the use of precipitation radar data to investigate the spatiotemporal effect of rainfall on freeways in Jacksonville, Florida. The linear regression analysis suggests a speed reduction of 0.75%, 1.54%, and 2.25% for light, moderate, and heavy rainfall, respectively. Additionally, headways were observed to increase by 0.26%, 0.54%, and 0.79% for light, moderate, and heavy rainfall, respectively. Measuring precipitation from radar data in lieu of using rain gauges has potential for improving the quality of weather data used for transportation engineering purposes. This approach addresses limitations experienced with conventional rain data, especially since conventional collection methods generally do not reflect the spatiotemporal distribution of rainfall.

Ramp metering is an effective management strategy, which helps to keep traffic density below the critical value, preventing breakdowns and thus maintaining the full capacity of the freeway. Warrants for ramp metering installation have been developed by a number of states around the nation. These warrants are generally simple and are based on the traffic, geometry, and safety conditions in the immediate vicinity of each ramp (local conditions). However, advanced applications of ramp metering utilize system-based metering algorithms that involve metering a number of on-ramps to address system bottleneck locations. These algorithms have been proven to perform better compared to local ramp metering algorithms. This has created a disconnection between existing agency metering warrants to install the meters and the subsequent management and operations of the ramp metering. Moreover, the existing local warrants only consider recurrent conditions to justify ramp metering installation with no consideration of the benefits of metering during non-recurrent events such as incidents and adverse weather. This dissertation proposed a methodology to identify the ramps to meter based on system-wide recurrent and non-recurrent traffic conditions. The methodology incorporates the stochastic nature of the demand and capacity and the impacts of incidents and weather using Monte Carlo simulation and a ramp selection procedure based on a linear programming formulation. The results of the Monte Carlo simulation are demand and capacity values that are used as inputs to the linear programming formulation to identify the ramps to be metered for each of the Monte Carlo experiments. This method allows the identification of the minimum number of ramps that need to be metered to keep the flows below capacities on the freeway mainline segment, while keeping the on-ramp queues from spilling back to the upstream arterial street segments. The methodology can be used in conjunction with the existing local warrants to identify the ramps that need to be metered. In addition, it can be used in benefit-cost analyses of ramp metering deployments and associated decisions, such as which ramps to meter and when to activate in real-time. The methodology is extended to address incidents and rainfall events, which result in non-recurrent congestion. For this purpose, the impacts of non-recurrent events on capacity and demand distributions are incorporated in the methodology.

Flow of traffic on freeways and limited access highways can be represented as a series of kinemetic waves. Solutions to these systems of equations become problematic under congested traffic flow conditions, and under complicated (real-world) networks. A simplified theory of kinematics waves was previously proposed. Simplifying elements includes translation of the problem to moving coordinate system, adoption of bi-linear speed-density relationships, and adoption of restrictive constraints at the on- and off-ramps. However, these simplifying assumptions preclude application of this technique to most practical situations. This research explores the limitations of the simplified theory of kinematic waves. First this research documents a relaxation of several key constraints. In the original theory, priority was given to on-ramp merging vehicles so that they can bypass any queue at the merge. This research proposes to relax this constraint using a capacity-based weighted fair queuing (CBWFQ) merge model. In the original theory, downstream queue affects upstream traffic as a whole and exiting traffic can always be able to leave as long as it gets to the diverge. This research proposes that this diverge constraint be replaced with a contribution-based weighted splitting (CBWS) diverge model. This research proposes a revised notation system, permitting the solution techniques to be extended to freeway networks with multiple freeways and their ramps. This research proposes a generalization to permit application of the revised theory to general transportation networks. A generalized CBWFQ merge model and a generalized CBWS diverge model are formulated to deal with merging and diverging traffic. Finally, this research presents computational procedure for solving the new system of equations. Comparisons of model predictions with field observations are conducted on GA 400 in Atlanta. Investigations into the performance of the proposed CBWFQ and CBWS models are conducted. Results are quite encouraging, quantitative measures suggest satisfactory accuracy with narrow confidence interval.

Made available in DSpace on 2014-06-11T19:28:27Z (GMT). No. of bitstreams: 0 Previous issue date: 2006-12-11Bitstream added on 2014-06-13T19:16:21Z : No. of bitstreams: 1 bordo_aa_me_prud.pdf: 2274588 bytes, checksum: e1175d5a0b4d8625ddf5307e742f634b (MD5) / Os eixos de desenvolvimento referem-se às grandes vias de circulação de pessoas, mercadorias, bens, serviços e informações, balizados por uma rede de centros urbanos, e que estimulam a localização das atividades industriais. No Estado de São Paulo, os eixos de desenvolvimento podem ser visualizados ao longo de importantes rodovias que fazem a ligação da metrópole paulista com o interior e que funcionaram como canais para a desconcentração urbano-industrial que se iniciou na década de 1970 e se intensificou na década de 1980. O eixo da rodovia Washington Luiz (trecho Araraquara São José do Rio Preto), duplicado no início da década de 1990, somente nos últimos anos passa a atuar como um dos fatores de localização industrial para o município de Itápolis na atração de algumas indústrias desconcentradas de outras regiões. No entanto, a maioria das indústrias de Itápolis é de capital local (70,37% das indústrias pesquisadas). Os principais ramos industriais observados na pesquisa foram as agroindústrias e as confecções. São indústrias classificadas como tradicionais ou de trabalho intensivo e estão ligadas aos antigos fatores de localização industrial, como a proximidade das matériasprimas e a disponibilidade de mão-de-obra. A economia de Itápolis é baseada fundamentalmente na agropecuária. O município possui o maior PIB agrícola do Brasil e foi responsável por 0,6% de toda a produção agrícola brasileira em 2002. Em 2003, a participação de Itápolis no valor adicionado da agricultura estadual foi de 1,8%. A citricultura é responsável pela geração de 80% do ICMS. O dinamismo da agricultura constitui-se em um importante atrativo para a atividade industrial, seja... / Development axles refer to big circulation routes of people, goods, services and information marked out by a network of urban centers, which foster industrial activities location. In the State of São Paulo, these development axles can be seen along important freeways that connect the capital with the country and they worked as channels for urban-industrial concentration changes, which began in the 70's and became stronger in the 80's. The axle of Washington Luiz freeway (between the cities of Araraquara and São José do Rio Preto), which was doubled at the beginning of the 90's, has only contributed to attract some industries from other areas to the district of Itápolis in the last few years. However, most of the industries in the town were born with local investment (70.37% of the ones researched). The main industrial branches observed were agriculture-industry and clothes. They can be considered traditional or intensive work and are linked to old factors of industrial location, such as being close to raw materials and work force availability. Itápolis economy is essentially based in agriculture. The district has the highest agricultural GDP of Brazil, being responsible for 0.6% of the Brazilian agricultural production in 2002. In 2003, its participation in the state agriculture was 1.8%. The citrus culture is responsible for producing 80% of ICMS (Brazilian tax over goods and services movement). A dynamic agriculture fully attracts industrial activity in the...(Complete abstract, click eletronic address below)