Machine learning is a powerful method to handle the self-driving problem. Researchers use machine learning to construct a neural network and train it to drive the car. A self-driving car is a safety-critical system. However, the neural network is not necessarily reliable. The output of a neural network can be easily influenced by many factors, such as the quality of training data and the runtime environment. Also, it takes time for the neural network to generate the output. That is, the self-driving car may not respond in time. Such weaknesses will increase the risk of accidents. In this thesis, considering the safety of self-driving cars, we apply a delay-aware shielding mechanism to the neural network to protect the self-driving car. Our approach is an improvement based on previous research on runtime safety enforcement for general cyber-physical systems that did not consider the delay to generate the output. Our approach contains two steps. The first is to use formal language to specify the safety properties of the system. The second step is to synthesize the specifications into a delay-aware enforcer called the shield, which enforces the violated output to satisfy the specifications during the whole delay. We use a lane keeping system as a small but representative case study to evaluate our approach. We utilize an end-to-end neural network as a typical implementation of such a lane keeping system. Our shield supervises those outputs of the neural network and verifies the safety properties during the whole delay period with a prediction. The shield can correct it if a violation exists. We use a 1/16 scale truck and construct a curvy lane to test our approach. We conduct the experiments both on a simulator and a real road to evaluate the performance of our proposed safety mechanism. The result shows the effectiveness of our approach. We improve the safety of a self-driving car and we will consider more comprehensive driving scenarios and safety features in the future. / Master of Science / Self-driving cars is a hot topic nowadays. Machine learning is a popular method to achieve self-driving cars. Machine learning constructs a neural network, which imitates a human driver's behavior to drive the car. However, a neural network is not necessarily reliable. Many things can mislead the neural network into making wrong decisions, such as insufficient training data or a complex driving environment. Thus, we need to guarantee the safety of self-driving cars. We are inspired to use formal language to specify the safety properties of the self-driving system. A system should always follow those specifications. Then the specifications are synthesized into an enforcer called the shield. When the system's output violates the specifications, the shield will modify the output to satisfy the specifications. Nevertheless, there is a problem with state-of-the-art research on specifications. When the specifications are synthesized into a shield, it does not consider the delay to compute the output. As a result, the specifications may not be always satisfied during the period of the delay. To solve such a problem, we propose a delay-aware shielding mechanism to continually protect the self-driving system. We use a lane keeping system as a small self-driving case study. We evaluate the effectiveness of our approach both on the simulation platform and the hardware platform. The experiments show that the safety of our self-driving car is enhanced. We intend to study more comprehensive driving scenarios and safety features in the future.

The focus of this thesis is the imposition of order onto a non-ordered space through the design of boundaries. The designed elements are derived from both the existing site conditions and the desire to improve the space, a threshold to the stadium. A building that acts as a boundary becomes a type of wall, having two very pragmatic sides. Therefore, in this case, these elements must also function as a filter. The challenge is to design a boundary or an edge condition which defines the site, in addition to framing the views and allowing people to pass through these thresholds. Likewise, circulation also helps to structure the site. The designed buildings allow for persons to pass along as well as to pass through. The elements placed onto the site allow for both formal and direct paths through the site, as well as more general and autonomous approaches to the stadium, located on the edge of the site. / Master of Architecture

Lane change or "Side" Crash Avoidance Systems (SCAS) technologies are becoming available to help alleviate the lane change crash problem. They detect lane change crash hazards and warn the driver of the presence of such hazards. This thesis examines driver lane change behavior and evaluates the potential effectiveness of five warning onset rules for lane change or "side" crash avoidance system (SCAS) technologies. The ideal SCAS should warn the driver only when two conditions are met: (1) positive indication of lane change intent and (2) positive detection of a proximal vehicle in the adjacent lane of concern. Together, these two conditions create a crash hazard. The development of SCAS technologies depends largely on an understanding of driver behavior and performance during lane change maneuvers. By quantifying lane change behavior, real world crash hazard scenarios can be simulated. This provides an opportunity to evaluate potential warning onset rules or algorithms of driver intent to change lanes. Five warning onset rules for SCAS were evaluated: turn-signal onset (TSO), minimum separation (MS), line crossing (LC), time-to-line crossing (TLC), and tolerance limit (TL). The effectiveness of each rule was measured by the maximum response time available (tavailable) to avoid a crash for a particular lane change crash scenario, and by the crash outcome, crashed or crash avoided, of a particular lane change crash scenario. / Master of Science

Real-time lane detection or localization is a crucial problem in modern Advanced Driver Assistance Systems (ADAS), especially in Automated Driving System. This thesis estimates the possibility to implement a lane detection system in the available low-power embedded hardware. Various state-of-the-art Lane Detection algorithms are assessed based on a number of proposed criteria. From the result of the evaluation, three different algorithms are constructed and implemented in the hardware using OpenCV library. The lane detection stage is done with different methods: using Hough Transform for line detection or randomly sampling hypotheses which are straight lines or cubic splines over the pre-processed binary image. Weights of the hypotheses are then calculated based on their positions in the image. The hypothesis which has highest weight or best position will be chosen to represent lane marking. To increase the performance of the system, tracking stage is introduced with the help of Particle Filter or Kalman Filter. The systems are then tested with several different datasets to evaluate the speed, performance and ability to work in real-time. In addition, the system interfaces with CAN bus using a CAN interface, so that the output data can be sent as messages via the CAN bus to other systems.

In Intelligent Transportation Systems, Advanced Vehicle Control and Safety System are one of the most important researches around the world. AVCSS is a technique applied on vehicle and is composed by sensor, computer, communication, and control. In order to keep driver safe, the technique covered Collision Avoidance, Longitudinal Automated Control, Lateral Automated Control, Automated Parking, etc, and Collision Avoidance, Longitudinal Automated Control, Lateral Automated Control are most important. This thesis implemented the Lateral Automated Control by using a CCD camera to extract the road environment. And I presented and analyzed lane detection about structured road and unstructured road. The structured road stands for its obvious lane mark such as general road and freeway; and the unstructured road stands for its unobvious lane mark or without lane mark such as country road and campus road. Because of the characteristic of lane mark, the structured road is easier to detect, and there were less research about unstructured road around the world. So this thesis focused on the unstructured lane detection, and implemented multi-system on DSP (Digital Signal Process). Finally, we applied intelligent control system to vehicle and successfully guided the vehicle in structure and unstructured road.

Thesis (Ph. D.)--Ohio State University, 1986. / Includes bibliographical references (leaves 150-154). Available online via OhioLINK's ETD Center.

Made available in DSpace on 2014-06-11T19:28:16Z (GMT). No. of bitstreams: 0 Previous issue date: 2011-10-13Bitstream added on 2014-06-13T19:36:38Z : No. of bitstreams: 1 santos_jv_me_mar.pdf: 1193000 bytes, checksum: fa96979567fc05011735cead8a9e0c2c (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / O presente trabalho tem como preocupação central mapear as contribuições de Horace Manley Lane na constituição da escola republicana paulista, entre 1890 a 1910. O foco será a ação de Horace Lane junto aos administradores da instrução pública paulista, em especial, Caetano de Campos e Oscar Thompson. Além da revisão dos trabalhos que focam a influência protestante na organização da instrução pública paulista, tomamos como fontes privilegiadas os relatórios oficiais, artigos editados em periódicos educacionais produzidos no período, documentos pessoais de Horace Lane e a edição do livro Education in the State of São Paulo, de autoria do Oscar Thompson, Horace Lane e Carlos Reis, apresentado na Feira Internacional de Saint Louis em 1904. O objetivo central é analisar atuação de Lane na instrução pública e evidenciar a influência da ação protestante em terras paulistas. Esta pesquisa está apoiada nas premissas de Marta Carvalho de uma História Cultural dos Saberes Pedagógicos, dialogando em torno, de suas práticas, seus agentes, suas representações e apropriações para uma nova compreensão da escola. Como resultado, este estudo permite dar visibilidade às contribuições de Horace Lane na consolidação da instrução pública paulista proposta pelos republicanos / The purpose of this study is to chart the contributions of Horce Manley in the constitution of the republican school of São Paulo, from around 1890 to 1910. The main focus is the actions of Horace Lane along the administrators of the public instruction of São Paulo, specially Caetano de Campos and Oscar Thompson. Besides the revision of studies that focus on the protestant influence in the organization of the public instruction of São Paulo, we used as privileged sources the official reports, personal documents of Horace Lane and the edition of the book „Eduaction in the State of São Paulo‟, whose authors are Oscar Thompson, Horace Lane and Carlos Reis. This book was presented at the Saint Louis International fair back in 1904. The main purpose is to analyze the interaction of Lane in the public instruction and to point the protestant influence of it. This research is supported in the assumptions of Marta Carvalho from the research “Cultural History of the pedagogical knowledge”, discussing around her practices, her agents, her representations and appropriations for a new comprehension of the school. As a result, this study allows us to highlight the contributions of Horace Lane in the consolidation of the public instruction of São Paulo from the Republicans

In dieser Arbeit wird das Verhalten der Fahrer vor, während und nach dem Spurwechsel anhand eines Datensatzes untersucht. Dieser wurde im Jahr 2006 in den USA erhoben und enthält die genauen Trajektoriendaten für jedes Fahrzeug. Es konnte gezeigt werden, dass schnellere Fahrzeuge ihren Spurwechsel auch schneller abschließen. Es ist auch erkennbar, dass bei einem Wechsel nach rechts kein kleinerer Abstand zum Vorderfahrzeug in Kauf genommen wird. Eine Untersuchung der freiwilligen und notwendigen Wechsel konnte keine signifikanten Unterschiede feststellen. Im zweiten Teil der Auswertung wurden die Spurwechselentscheidungen mittels eines generellen Modells und MOBIL untersucht. Die einzusetzenden Werte wurden mit den Fahrzeugfolgemodellen OVM, Gipps-Modell und IDM erzeugt. Die Entscheidungen der Spurwechselmodelle fielen nur selten so aus, wie sie in der Realität getroffen wurden. Damit werden die Modelle ihrem eigenen Anspruch nicht gerecht, die Realität möglichst gut nachzubilden.:1 Einleitung 1 2 Theoretische Grundlagen 3 2.1 Begriffserklärungen 3 2.2 Fahrstreifenwechsel 6 2.2.1 Formen von Spurwechseln 6 2.2.2 Ebenen von Entscheidungen 7 2.3 Verkehrsmodelle 8 2.3.1 Mathematische Grundlagen 9 2.3.2 Fahrzeugfolgemodelle 9 2.3.3 Simulation des Fahrstreifenwechsels 12 3 Aufbereitung des Datensatzes 15 3.1 Untersuchungsgebiet 15 3.2 Datensatz 16 3.3 Datenaufbereitung 18 3.3.1 Bestimmung der lokalen Geschwindigkeit 18 3.3.2 Bestimmung der lokalen Beschleunigung 21 3.3.3 Bestimmung der Abstände und Zeitlücken 22 3.4 Bekannte Fehler im Datensatz 23 3.5 Implementierung und Kalibrierung der Modelle 28 4 Auswertung des Datensatzes 31 4.1 Ermittlung wichtiger Kenngrößen 31 4.2 Deskriptive Beschreibung von Fahrstreifenwechseln 33 4.2.1 Freiwillige Spurwechsel 33 4.2.2 Notwendige Spurwechsel 37 4.3 Modellbasierte Auswertung 39 4.3.1 Kalibrierung und Validierung 39 4.3.2 Untersuchung der Spurwechselentscheidung 41 4.3.3 Untersuchung der Spurwechselentscheidung mit MOBIL 43 5 Zusammenfassung und Ausblick 47 5.1 Zusammenfassung und Überprüfung der Hypothesen 47 5.2 Weitere Forschungsmöglichkeiten 50 / The topic of this master thesis is the investigation of the behaviour of the drivers before, during and after lane changes on highways. For this purpose a data set of the I-80 in the USA was used. It was collected in 2006 and contains the trajectories of all vehicles. It could be shown that faster cars even do a faster lane change. The drivers also do not accept a smaller gap to the preceding car after a lane change to the right. A comparison of voluntary and necessary lane changes did not show significant differences. In the second part of the analysis the lane change decision was examined using a general model and MOBIL. The data to be inserted were generated using the car-following models OVM, Gipps-model and IDM. But the outcomes of the lane change models only rarely matched reality. Therefore, the models did not fulfill their own standards to picture the existing reality.:1 Einleitung 1 2 Theoretische Grundlagen 3 2.1 Begriffserklärungen 3 2.2 Fahrstreifenwechsel 6 2.2.1 Formen von Spurwechseln 6 2.2.2 Ebenen von Entscheidungen 7 2.3 Verkehrsmodelle 8 2.3.1 Mathematische Grundlagen 9 2.3.2 Fahrzeugfolgemodelle 9 2.3.3 Simulation des Fahrstreifenwechsels 12 3 Aufbereitung des Datensatzes 15 3.1 Untersuchungsgebiet 15 3.2 Datensatz 16 3.3 Datenaufbereitung 18 3.3.1 Bestimmung der lokalen Geschwindigkeit 18 3.3.2 Bestimmung der lokalen Beschleunigung 21 3.3.3 Bestimmung der Abstände und Zeitlücken 22 3.4 Bekannte Fehler im Datensatz 23 3.5 Implementierung und Kalibrierung der Modelle 28 4 Auswertung des Datensatzes 31 4.1 Ermittlung wichtiger Kenngrößen 31 4.2 Deskriptive Beschreibung von Fahrstreifenwechseln 33 4.2.1 Freiwillige Spurwechsel 33 4.2.2 Notwendige Spurwechsel 37 4.3 Modellbasierte Auswertung 39 4.3.1 Kalibrierung und Validierung 39 4.3.2 Untersuchung der Spurwechselentscheidung 41 4.3.3 Untersuchung der Spurwechselentscheidung mit MOBIL 43 5 Zusammenfassung und Ausblick 47 5.1 Zusammenfassung und Überprüfung der Hypothesen 47 5.2 Weitere Forschungsmöglichkeiten 50

Recent years have seen a renewed interest in Variable Speed Limit (VSL) strategies. New opportunities for VSL as a freeway metering mechanism or a homogenization scheme to reduce speed differences and lane changing maneuvers are being explored. This paper examines both the macroscopic and microscopic effects of different speed limits on a traffic stream, especially when adopting low speed limits. To that end, data from a VSL experiment carried out on a freeway in Spain are used. Data include vehicle counts, speeds and occupancy per lane, as well as lane changing rates for three days, each with a different fixed speed limit (80 km/h, 60 km/h, and 40 km/h). Results reveal some of the mechanisms through which VSL affects traffic performance, specifically the flow and speed distribution across lanes, as well as the ensuing lane changing maneuvers. It is confirmed that the lower the speed limit, the higher the occupancy to achieve a given flow. This result has been observed even for relatively high flows and low speed limits. For instance, a stable flow of 1942 veh/h/lane has been measured with the 40 km/h speed limit in force. The corresponding occupancy was 33%, doubling the typical occupancy for this flow in the absence of speed limits. This means that VSL strategies aiming to restrict the mainline flow on a freeway by using low speed limits will need to be applied carefully, avoiding conditions as the ones presented here, where speed limits have a reduced ability to limit flows. On the other hand, VSL strategies trying to get the most from the increased vehicle storage capacity of freeways under low speed limits might be rather promising. Additionally, results show that lower speed limits increase the speed differences across lanes for moderate demands. This, in turn, also increases the lane changing rate. This means that VSL strategies aiming to homogenize traffic and reduce lane changing activity might not be successful when adopting such low speed limits. In contrast, lower speed limits widen the range of flows under uniform lane flow distributions, so that, even for moderate to low demands, the under-utilization of any lane is avoided. These findings are useful for the development of better traffic models that are able to emulate these effects. Moreover, they are crucial for the implementation and assessment of VSL strategies and other traffic control algorithms.

The implementation of managed lanes (MLs), also known as dynamically priced express lanes, to improve freeway traffic flow and personal throughput is on the rise. Congestion pricing is increasingly becoming a common strategy for congestion management, often requiring microscopic simulation during both planning and operational stages. VISSIM is a recognized microscopic simulation software used for analyzing the performance of managed lanes (MLs). This thesis addressed two important microscopic simulation issues that affect the evaluation results of MLs. One of the microscopic simulation issues that has not yet been addressed by previous studies is the required minimum managed lane routing decision (MLRD) distance upstream of the ingress point of MLs. Decision distance is an optimal upstream distance prior to the ingress at which drivers decide to use MLs and change lanes to orient on a side of MLs ingress. To answer this question, this study used a VISSIM model simulating I-295 proposed MLs in Jacksonville, Florida, United States (U.S), varying the MLRD point at regular intervals from 500 feet to 7,000 feet for different levels of service (LOS) input. Three measures of effectiveness (MOEs) - speed, the number of vehicles changing lanes, and following distance - were used for the analysis. These MOEs were measured in the 500 feet zone prior to the ingress. The results indicate that as the LOS deteriorates, speed decreases, the number of vehicles changing lanes increases, and the following distance decreases. When the LOS is constant, the increase in the MLRD distance from the ingress point was associated with the increase in the speed at the 500 feet zone prior to the ingress, less number of lane changes, and the increase in following vehicle gap. However, the MOEs approached constant values after reaching a certain MLRD distance. LOS D was used to determine the minimum MLRD distance to the ingress of the MLs. The determined minimum MLRD distances were 4,000 and 3,000 feet for 6 and 3 lane segments prior to the ingress point, respectively. Another issue addressed in this thesis is the managed lane evaluation (MLE) outputs, which include speed, travel time, density, and tolls. In computing the performance measures, the existing VISSIM managed lane evaluation (EVMLE) tool is designed to use the section starting at the point when vehicles are assigned to use MLs, also known as the MLRD point, which is located upstream of the ingress. The longer the MLRD distance from the ingress, the more the EVMLE tool uses the traffic conditions of the MLs traffic before entering the ML in its computations. This study evaluates the impact of the MLRD distance on the EVMLE outputs and presents a proposed algorithm that addresses the EVMLE shortcomings. In order to examine the influence of the MLRD distance on the outputs of the above-mentioned two algorithms, simulation scenarios of varying MLRD distances from 500 ft to 7,000 feet from the ingress were created. For demonstration purposes, only the speed was used to represent other performance measures. The analysis of variance (ANOVA) test was performed to determine whether there was a significant difference in the speed results with the change in the MLRD distance. According to the ANOVA results, the EVMLE tool produced ML speeds that are MLRD dependent, yielding lower speeds with an increased MLRD distance. On the other hand, the ML speed results from the proposed algorithm were fairly constant, regardless of the MLRD distance.