Handbook of Intelligent Vehicles

Intelligent vehicles can make road traffic safer, more efficient, and cleaner. We provide an overview of current and emerging intelligent functions in vehicles. We focus on intelligent functions that actively interfere with the task of driving a vehicle. We give a classification based on the driving task, the type of road, and the level of support. We give a review of route guidance systems, advanced driver assistance systems, as well as automated vehicles. We discuss the non-technological issues that need to be addressed for the successful deployment of intelligent vehicles, such as cooperation between industry and public authorities, increasing awareness amongst stakeholders, research and development, and legal framework. Finally, we introduce the subjects that will be addressed in the remainder of this section.

Nowadays, cars are equipped with more electronic systems than in the past. Today, vehicles are equipped with hundreds of miniature sensing systems, such as temperature, tire pressure, accelerometer, and speed sensors. Actuators are also installed as components of advanced systems for optimized braking and assisted steering, although the market for really intervening systems is not mature yet.In this chapter, different sensing and actuation systems are highlighted. The purpose of this chapter is to give an overview of these systems and do not go deep into detail about the different technologies behind such systems. The sensors are grouped into three different categories: general in-vehicle sensors, perception sensors, and virtual sensors. Most of the general in-vehicle sensors are already available in the automotive market in the majority of commercial cars. On the other hand, the market penetration rate of perception sensors, except for ultrasonic sensors, is very low mainly because of their cost. Finally, there are some information sources that are not actual sensors and play a significant role in automotive applications, such as the digital maps. Actuators are first distinguished and described according to their energy source into mechanical actuators, electrical actuators, pneumatic or hydraulic actuators, piezoelectric actuators, and thermal bimorphs. Next, the design of advanced intervening systems is presented, namely, the ABS, electronic stability control, autonomous cruise control, assisted steering. More advanced systems like the steer by wire and brake by wire are also presented, although they have not yet entered the market as products, with few exceptions. Finally, the vision of a fully automated vehicle is presented together with the considerations that still accompany it, and some first prototypes and research work toward this direction are highlighted.

Cooperative intelligent vehicle systems constitute a promising way to improving traffic throughput, safety and comfort, and thus are the focus of intensive research and development. The vehicles implement more and more complex onboard functionalities, which interact with each other and with their surroundings, including other vehicles and roadside information infrastructure. In order for a functionality “do the right thing” it should have a sufficiently complete and certain interpretation of the surrounding world (i.e., relevant part of the road infrastructure, the surrounding vehicles, the ego-vehicle itself, etc.). Due to the ever-existing limitations of the sensing, the complexity of the data interpretation and the inherent uncertainty of the world “out there”, creating this representation poses major challenges and has far reaching consequences concerning how onboard functionalities should be built. The situational awareness term covers an overarching research field, which addresses this understanding process from different angles. It attempts the conceptualization of the problem domain, it relates sensory data processing and data fusion with the understanding process, and it investigates the role of humans in the related processes and even gives architectural guidelines for system design.First a brief overview is given about the established models for situational awareness emphasizing the specialties of the intelligent vehicle systems. Then the representation problem is covered in details because the representation has strong influence both on the sensing, data interpretation, control and architectural aspect. Finally the control and architectural aspects are covered addressing the design for dependability.

We present a survey on traffic management and control frameworks for Intelligent Vehicle Highway Systems (IVHS). First, we give a short overview of the main currently used traffic control methods that can be applied in IVHS. Next, various traffic management architectures for IVHS such as PATH, Dolphin, Auto21 CDS, etc., are briefly discussed and a comparison of the various frameworks is presented. Subsequently, we focus on control of vehicles inside a platoon, and we present a detailed discussion on the notion of string stability. Next, we consider higher-level control of platoons of vehicles. Finally, we present an outlook on open problems and topics for future research.

One purpose of Intelligent Vehicles is to improve road safety, throughput, and emissions. However, the predicted effects are not always as large as aimed for. Part of this is due to indirect behavioral changes of drivers, also called behavioral adaptation. Behavioral adaptation (BA) refers to unintended behavior that arises following a change to the road traffic system. Qualitative models of behavioral adaptation (formerly known as risk compensation) describe BA by the change in the subjectively perceived enhancement of the safety margins. If a driver thinks that the system is able to enhance safety and also perceives the change in behavior as advantageous, adaptation occurs. The amount of adaptation is (indirectly) influenced by the driver personality and trust in the system. This also means that the amount of adaptation differs between user groups and even within one driver or changes over time.Examples of behavioral change are the generation of extra mobility (e.g., taking the car instead of the train), road use by “less qualified” drivers (e.g., novice drivers), driving under more difficult conditions (e.g., driving on slippery roads), or a change in distance to the vehicle ahead (e.g., driving closer to a lead vehicle with ABS).In effect predictions, behavioral adaptation should be taken into account. Even though it may reduce beneficial effects, BA (normally) does not eliminate the positive effects. How much the effects are reduced depends on the type of ADAS, the design of the ADAS, the driver, the current state of the driver, and the local traffic and weather conditions.

The development of systems for intelligent functions is a very complex process, involving many technological components, requirements with respect to robustness and fail safety, time pressure, and different stakeholders. Simulation plays an important role in the development process and can support the development process from the early concept development, through testing and verification of components to the validation and evaluation of systems. Requirements and examples of simulation environments needed are discussed. Traffic flow simulation is used to show how it can be used in the development of the concept of Cooperative Adaptive Cruise Control. Next, simulation is applied for testing and verification of cooperative safety applications. Finally, simulation is applied for validation and evaluation of applications tested in a Field Operational Test.

Decades of research efforts have greatly advanced our understanding of vehicle longitudinal control and yielded fruitful results. This chapter provides a detailed discussion on this integral part of vehicle regulation control. This chapter starts with an introduction that defines the scope of our discussion as the longitudinal control of automated vehicles, provides the relevant research history, and describes the functionality of vehicle longitudinal control. Subsequently, the system requirements and framework design are discussed. As a safety-critical system, the longitudinal control of automated vehicles needs to satisfy both safety requirements and performance requirements. Typically formulated as a feedback control system, the longitudinal control system consists of sensors (and sensor processing), control computation, and control actuation components. This chapter further describes the longitudinal control systems for passenger vehicles in detail, which covers the sensing, modeling, and controller design by using a longitudinal control system designed for automated vehicles in a platoon as an example. To further extend the discussion to automated heavy vehicles, a specific control application, precision-stopping control for automated buses, is discussed. This chapter concludes with a short summary of the current status and thoughts on future directions for the longitudinal control of automated vehicles.

The adaptive cruise control (ACC) and cooperative adaptive cruise control (CACC) system is the extension to the conventional cruise control (CC). This chapter focuses on the introduction of various design methodologies for ACC/CACC controllers. The ACC/CACC operation-mode transition and system architecture are presented in detail together with different system components to illustrate concepts and functions of ACC and CACC. A unified control problem formulation and multiple design objectives are then described. Different control design methodologies such as linear control design, nonlinear control design, model predictive control design, and fuzzy control design are reviewed.

Run-off-road crashes are responsible for about 34% of road fatalities. In addition, side swiping crashes are also caused by either unintended or ill-timed lane departures. Many systems have been developed to reduce these kinds of accidents. Vehicle lateral control and warning systems are introduced for this reason. Vehicle lateral control systems can help the drivers to change their lanes safely or assist them in parking or even performing evasive maneuvers. This chapter aims to serve as an introduction to dynamics, control, and modeling of vehicle lateral motion in the context of intelligent vehicles. First, the major components that determine lateral dynamics of the vehicle are briefly described. Then, the basics of tire dynamics and vehicle modeling are presented. Many different linear and nonlinear controllers have been suggested in the literature to control the combined lateral and longitudinal motion of the vehicle. It is expected that in the future vehicle lateral control will be extended to performing lane change or evasive maneuvers. A brief review of the control methods and some of the experimental implementations of autonomous trajectory following (lane change maneuvers) are presented. A few examples of their commercial productions are provided. Finally, some concluding remarks about the future of vehicle lateral control and the current state of the development in this area are given.

Competitiveness to a company is given by innovations. The chassis as main part in vehicle design is incisive to the driving behavior of a car. On the one side, mechanical devices are well-engineered which means differentiation to competitors in mechanical devices is complex and costly. On the other side, improvements due to clients and legislator such as driving dynamics, CO₂ reduction, or pedestrian protection increase the requirements to the chassis concerning comfort, safety, handling, or individualization by less cost and maintenance.This balancing act can be done by mechatronics systems which means the interaction of mechanic, electronic, and informatics devices. Basic mechatronics systems are used to assist the driver (e.g., power steering) or to overrule a wrong driver input (e.g., ABS brake). Different from this so-called by-wire systems are extensive mechatronics systems where the vehicle behavior and the driver feedback can be designed independently (there is no mechanical link between input and output).Drive-by-wire, X-by-wire, or simply by-wire technology is already present nowadays. Starting with aeronautics, where fly-by-wire has been used extensively in the Airbus A320 family without mechanical backup. In passenger cars, by-wire functionality and by-wire systems are far more recent, but still already well known (VDI-Bericht 1828, 2004).One can distinguish between by-wire functionality and by-wire system. The by-wire functionality can be reduced to the ability to control or even only apply a force by an electrical signal (through an electrical wire) to the vehicle. The definition by-wire system is that the line between the driver’s input interface and the actuation which produces force is partly designed by wire. Hence, in contrast to the by-wire functionality, the system has no permanent hydraulic or mechanic linkage between them.Common advantages of by-wire systems are the freedom in functionality, package integration, reducing variants, design, and enabler for driver assistance functions. In Sect. 1, these general facts of by-wire systems, the vehicle-driver control loop, and aspects of the input module behavior will be shown. Afterward longitudinal and lateral dynamic systems and their functionality are explained in more detail (Sects. 2 and 3).One benefit of by-wire systems is the system and functional integration. In Sect. 4 integrated corner modules are illustrated and analyzed as well as integrated control strategy aspects. The challenge of by-wire systems are the functional safety requirements, especially in terms of availability. The latter is the most important for the OEMs and customers. Any minor failure of the systems, which has to be displayed in the dash board, will reduce the customers’ faith in the car. These aspects will be explained in Sect. 5.

This chapter focuses on a methodological comparison of the energy consumption of passenger cars with conventional internal combustion engine and new electric powertrains.The scientific methodology of the presented knowledge relies on simulation models validated by real-world tank-to-wheel emission and energy measurements performed in urban and extra urban areas as well as on freeways.Furthermore, an overview of the recent development of energy storage systems, energy converters, and powertrains in developed regions is given.Based on a statistically average European passenger car, fuel and energy consumption of conventional diesel, gasoline hybrid, fuel cell, and battery electric powertrains are compared. Special focus is given to the real-world energy consumption of the vehicles. This means auxiliary power units of the car such as onboard electronics, safety systems, heating, and air conditioning have significant influence on the fuel consumption, especially in driving conditions with lower speeds.The results show that for typical real-world operational conditions during the seasons of a whole year the hybrid vehicle and, with some limitations in the range of operation, fuel cell vehicle are comparable to the usability of a conventional vehicle. Battery electric vehicles (BEV) have very limited range.For the consistent comparison of propulsion systems with such different energy carriers, like liquid fuels, gaseous hydrogen, and electric power, the energy consumption with respect to primary energy supply was carried out. Based on this approach, the distance related energy consumption of hybrid and fuel cell vehicles is somewhat favorable compared to the conventional vehicle. Battery electric vehicles consume significantly more energy due to inefficiencies in power production and transmission as well as charging and discharging losses.Nevertheless, cars with fuel cell and battery electric propulsion systems have the considerable benefit of local emissions-free driving.

The emergence of global navigation satellite systems (GNSSs) has enabled tremendous development in vehicular navigation for various applications. The GNSS technology provides a unique global positioning capability with meter-level accuracy at a low hardware cost and zero marginal infrastructure cost. The GNSSs work by using the satellites as radio beacons, broadcasting a satellite-specific signal and their own position. The range to the satellites is measured up to a common clock offset, and any user equipped with a GNSS receiver capable of receiving the signal from four or more satellites can position itself by multilateration. The position, being a fundamental piece of information for automatizing and facilitating location-dependent system interaction and services, makes the GNSS an enabling technology for many intelligent vehicle and transportation system capabilities.This chapter will focus on introducing the basic principles of the GNSS technology and the signal processing that allows the GNSS receiver to determine its position: in Sect. 1, the technology, its limitations, and currently available GNSSs are reviewed; in Sect. 2, the principles of the GNSS positioning, the signal characteristics, and fundamental components are discussed; in Sect. 3, the theoretical relations governing the positioning are presented; in Sect. 4, implementation-related issues, error sources, and GNSS receivers are discussed; in Sect. 5, the method of differential GNSS is introduced and current augmentation systems are reviewed; and finally in Sect. 6, conclusions are drawn and references, for further reading about different aspects of the GNSS technology, are given.

In-vehicle navigation systems usually rely on the integration of data from a range of positioning sensors/systems such as GPS or GPS integrated with other positioning sensors. Even with very robust sensor calibration and sensor fusion methods, positioning inaccuracies are sometimes unavoidable. In addition, there are inaccuracies with a digital road map due to errors in map creation, projection and digitization. As a result of such imprecision in the positioning systems and the faulty digital base map, actual vehicle positions do not always match with the spatial road map although the vehicle is known to be restricted on the road network. This phenomenon is referred to as spatial mismatch. The spatial mismatch is often more severe at junctions, roundabouts, complicated flyovers and built-up urban areas with complex route structures. However, an intelligent algorithm can be formulated by taking into account the historical trajectory of the vehicle and topological information of the road network (e.g., connectivity and orientation of links) to precisely identify the correct link on which a vehicle is traveling. Furthermore, an estimation of the vehicle location on the link can also be determined by taking into account all error sources associated with the positioning systems and digital map database. This is known as a map-matching algorithm. This chapter discusses the considerable momentum in research and development activities in map-matching, especially map data quality, methods and reliability issues surrounding map-matching algorithms. Future developments of map-matching algorithms and how such algorithms can tackle the positioning and navigation requirements of autonomous navigation are also discussed.

Situational awareness is of paramount importance in all advanced driver assistance systems. Situational awareness can be split into the tasks of tracking moving vehicles and mapping stationary objects in the immediate surroundings of the vehicle as it moves. This chapter focuses on the map estimation problem. The map is constructed from sensor measurements from radars, lasers and/or cameras, with support from on-board sensors for compensating for the ego-motion.Four different types of maps are discussed:(i)Feature-based maps are represented by a set of salient features, such as tree trunks, corners of buildings, lampposts and traffic signs.(ii)Road maps make use of the fact that roads are highly structured, since they are built according to clearly specified road construction standards. This allows relatively simple and powerful models of the road to be employed.(iii)Location-based maps consist of a grid, where the value of each element describes the property of the specific coordinate.(iv)Finally, intensity-based maps can be considered as a continuous version of the location-based maps.The aim is to provide a self-contained presentation of how these maps can be built from measurements. Real data from Swedish roads are used throughout the chapter to illustrate the methods.

The performance of all navigation and tracking algorithms for road-bound vehicles can be improved by utilizing the trajectory constraint imposed from the road network. We refer to this approach as road-assisted navigation and tracking. Further, we refer to the process of incorporating the road constraint into the standard filter algorithms by dynamic map matching. Basically, dynamic map matching can be done in three different ways: (1) as a virtual measurement, (2) as a state noise constraint, or (3) as a manifold estimation problem where the state space is reduced. Besides this basic choice of approach, we survey the field from various perspectives: which filter that is applied, which dynamic model that is used to describe the motion of the vehicle, and which sensors that are used and their corresponding sensor models. Various applications using real data are presented as illustrations.

The basics around in-car navigation is discussed, including the principals of contemporary systems, global navigation satellite system basics, dead-reckoning, map-matching, and strategies for information fusion. In-car navigation system are generally made out of three building blocks, an information source block, an information fusion block, and an user interface block. This chapter presents an overview of the information source block and the information fusion block. First, the ideas of operation and main characteristics of the four most commonly used information sources, global navigation satellite systems, vehicle motion sensors, road maps, and mathematical models of the vehicle dynamics, are reviewed. Thereafter, common techniques to combine the information from the different information sources into an estimate of the position, velocity, etc. of the car are reviewed.

In-car navigation systems do not consist of only navigation functions that are combinations of GPS and Map. Present in-car navigation systems are an integrated system that mainly consists of navigation function, Audio and Video function, and communication function. This chapter provides the introduction of in-car navigation system that has various functions and connects different devices, and also shows future in-car navigation systems. Firstly, this chapter provides the basic knowledge of in-car navigation system by tracing back through the history of in-car navigation system and the system architecture. These explanations give insight into the main hardware and software components of in-car navigation systems. Further, explaining the architecture of Portable Navigation Device (PND), it shows the characteristic of PND. Secondly, this chapter briefly describes, showing trends for the future, the main software components of in-car navigation systems, i.e., navigation function, audio and video function, and communication function. Enhancement of navigation functions seems to be slowing down, but there is still considerable room for growth by linking to network. Voice recognition and speech synthesis, also covered in this chapter, would become more attractive function by linking to network. In-car navigation systems are designed to connect various devices, e.g., smart phone, portable audio device, camera, rear monitor, and ITS devices. This chapter also describes about two more devices: camera device to be connected to in-car navigation system and display device that takes center stage of front. Lastly, we look “green technology” application of navigation functions, and the in-car navigation system for electric vehicles, which functions would be different from previous in-car navigation system to provide the useful applications.

Driving is a complex task of strategic decision making, maneuvering and controlling the vehicle while responding to external stimuli, traffic laws, and imminent hazards. Driver’s cognitive perception and reaction, physiological, and psychological capabilities along with experience, age, and many other factors play a major role in shaping the driving behavior and the skills to control the vehicle. Driver assistance systems are designed to support the driver in performing the primary driving tasks and the secondary in-vehicle tasks that may be required (operating radio, etc.). The goal of driver assistance or ADAS (advanced driver assistance systems) is to enhance safety, comfort, and efficiency of driving by intervening in the handling aspects of the vehicle and supporting the secondary tasks for comfort, navigation, etc. Driver assistance deals with the environment in terms of sensing and responding, the vehicle in terms of sensing and actuating electromechanical systems, and most importantly the driver in terms of augmenting information, enhancing sensing capabilities, and assisting in control functions. This chapter examines various aspects of driver assistance system including driver cognitive perception–response, system types and classifications, integrated safety, man–machine interface, and evaluation of effectiveness. This chapter concludes with listing existing ADAS and research needs.

In this chapter, the author presents a general framework classifying the different models adopted for capturing driver behavior focusing on the human cognitive dimensions and the traffic decision-making dimensions. Special interest is directed toward the “lower-level” microscopic models that can be linked directly to two core driving assistance technologies: adaptive cruise controls and lane-departure warning systems. These “lower-level” models are classified either as acceleration models or as lane changing models.Acceleration models are at the core of operational driving behaviors, and include car-following models which capture interactions between a lead vehicle and following vehicles. The main assumption in these models is that the behavior of the following vehicle is directly related to a stimulus observed/perceived by the driver, defined relative to the lead vehicle. In addition to the operational aspect, lane changing models capture the tactical side of driving. Most lane changing models have followed a deterministic rule-based framework where changing lanes is directly related to the desirability of such maneuver, its necessity, and its possibility/safety. Recognizing the limitations of the major existing microscopic traffic models, the objective in this chapter is to advance the state of knowledge in modeling driver behavioral processes and to offer an insight into current modeling approaches and the corresponding advantages and disadvantages.

Naturalistic driving research is the in situ investigation of driver performance and behavior. Video cameras and a suite of sensors are installed on participants own vehicles and are used to continuously record the driver, the vehicle, and the environment over an extended period of time. The collected data typically span hundreds of thousands of vehicle-miles-traveled and provide an “instant replay” of the rare occurrence of safety-critical events. The method supports the representative design of experiments, where the drivers, vehicles, and environment sampled are representative of the conditions to which the results are applied. Naturalistic Driving Studies (NDS) are an effective tool for the design, testing, and evaluation of driver assistance systems. This is because they can support various stages of a user-center systems design process. First, NDSs can help determine what drivers need from a new driver assistance system by allowing researchers to assess the driver error contributing to safety-critical events. Secondly, the approach can serve the testing of working prototypes, where “natural” driver behavior and performance with the candidate driver assistance systems is observed. Thirdly, novel test criteria, such as drivers’ rate of involvement in safety-critical events, can be used to evaluate the driver assistance systems’ effectiveness at improving driver performance. NDSs and their role in the design, testing, and evaluation of driver assistance systems are described in this chapter.

Intelligent Speed Adaptation (ISA) systems are in-vehicle systems designed to improve driver compliance with safe speeds. These systems can provide information on safe speeds to driver, warn the driver when they are exceeding this limit, or control brakes or throttle to prevent speeding. Because of the link between excessive speeding and severe crashes, ISA systems have been called “the most powerful collision avoidance system currently available” (Carsten and Tate 2001). However, ISA systems do face challenges to their widespread deployment. Perhaps the most significant of these challenges is finding an appropriate balance between user acceptability and system effectiveness. The more effective that an ISA system is at reducing speeding, the less likely it is to be acceptable to drivers, particularly those who would benefit the most from ISA systems. In addition, some researchers have expressed concern about potential negative safety implications of ISA systems including driver unloading, driver distraction, negative behavioral adaptations, and negative interactions with other road users. This chapter presents an overview of ISA system configurations, potential benefits of ISA systems, and challenges faced by the systems. In addition, case studies, including large-scale field tests of ISA systems, are presented.

In recent years, research and development activities of automobile manufacturers have placed an increasing focus on offering intelligent assistance systems in the vehicle. By delivering targeted information and warnings, by delegation of tasks, or by intervention, these functions aim to improve active safety, particularly in complex situations, and/or to enhance the driver’s sense of comfort. The system of driver, vehicle, and environment can be thought of as a control loop with feedback, in which the role of human drivers is decisive in determining the safety potential of this control loop. The safety and comfort characteristics arise for the driver from his interactions with the vehicle and the environment (Bernotat 1970). The human–machine interaction serves as the interface between the driver and the vehicle and also between the driver and the environment. Thus, the design of the human–machine interface is a key determinant of the effectiveness and acceptance of driver assistance systems. Due to the increasing amount of systems in a car, the functional integration of different assistance systems and a higher degree of automation of the functions are expected in future.

Adaptive Cruise Control (ACC) has reached a new quality in driver assistance. For the first time, a large part of the driver’s tasks can be assigned to an automatic system and the driver relieved to a substantial degree. Based on Cruise Control, ACC adjusts the vehicle speed to the surrounding traffic. It accelerates and decelerates automatically when a preceding vehicle is traveling at less than the speed desired by the driver.ACC is a key functional innovation and represents a new system architecture with a high degree of function distribution. The different operating modes and system states are described along with function limits and transition conditions.From the many elements of this overall function, target selection and longitudinal control are addressed in detail because of the special challenges they present. Target selection is based on the actual road curvature being determined by the ESC sensor signals that describe the driving dynamics, of which several options are assessed. Predicting and selecting a suitably shaped corridor is explained using an example. Major sources of error for the individual steps, their severity, and possible countermeasures are described.The prerequisite for vehicle-following-distance control is the selection of a target. An example shows that the basic control principle is simple, but it conflicts with comfort and convoy stability. Details of additional control functions in curve situations and approaches are provided.The driver perspective is addressed in terms of control and display functions and in terms of satisfaction as ascertained by use and acceptance studies, also taking into account an extended driver familiarization phase.

Forward collisions represent a significant portion of all severe accidents. This is why appropriate warning and collision avoidance systems are of great importance to increase traffic safety. Different system specifications are subsumed under the so-called FCX-systems; they differ in their way of affecting the overall system driver–vehicle–environment as forward collision conditioning, forward collision mitigation, forward collision warning, and forward collision avoidance systems. A more specific definition of FCX-systems is derived by distinguishing them from other related systems as adaptive cruise control and pedestrian safety systems which can also have an impact on forward collisions. The specifications of the actual systems already on the market can only be understood if the characteristics of machine perception are considered carefully. The progress in the field of machine perception enables the forward collision warning and avoidance systems. There are still limitations of state-of-the-art perception systems compared to attentive human drivers which must be considered when designing FCX-functions. The state of the art in FCX-systems is sketched highlighting realized examples of FCX-systems of different car manufacturers. The last focus is on a systematic design process which is recommended for driver assistance systems. The motivation for the assistance is always to be derived from accident research. Already in the early conceptual phase functional safety, legal, ergonomic, and marketing aspects should be taken into consideration. Only if a consistent functional specification is found, further developments including package and architecture aspects are justified. Concepts for testing and evaluation should be designed in an early development phase as well.

Supporting the driver for keeping the lane is the aim of lane departure warning and lane keeping systems. Unintended lane departures account for a high percentage of road accidents in general and especially of severe accidents. Therefore, these systems have been researched intensively and introduced in the market for several vehicles from heavy trucks to compact sedans. The systems described in this chapter aim at increasing the safety of driving or increasing the comfort on long journeys or a combination of both. The characteristics of the different variants are described as well as key technologies of state-of-the-art implementations.

In developed countries such as the USA or Europe, the risks of injury or fatality in traffic accidents have declined significantly in recent years. These reductions apply to both vehicle passengers and other involved persons. Much of this improvement has been attributable to progress in the field of passive safety, i.e., better protection of car occupants in situations where an accident is unavoidable. However, the marginal benefits resulting from additional efforts and expenditures in passive safety have begun to decrease; in other words, a classical “point of diminishing returns” has been reached. Increasing emphasis for achieving further significant improvements in vehicle safety will be placed on integral safety systems: Integral safety involves a concerted strategy of interlinking sensors and actuators of active and passive safety. The primary goal of this interlinking is optimization of performance and robustness of safety systems for occupants, but integral safety approaches can also achieve better protection of vulnerable road users than passive safety measures alone. In view of considerations such as reduction of CO₂ and fuel consumption, there is another attractive benefit: integral safety can serve to reduce the steady weight increase of vehicles and thus provide an important contribution to the development of both sustainable and safe vehicles.In order to develop effective measures for mitigating the severity of traffic accidents or even completely avoiding them, it is essential to understand the mechanisms of accident events, including the processes and risks involved in traffic situations in which these accidents occur. A quantitative understanding of these processes and risks aids in assessing the potential effectiveness of vehicle safety measures. The automobile industry is faced with enormous challenges in discovering and implementing the most effective solutions. Assessment by legal authorities and/or consumer groups should concentrate on safety performance, not on specification of particular technologies or methodologies, and should encourage implementation of devices providing greatest safety benefits by mandating robust and standardized testing and assessment techniques that quantify and measure effectiveness independently of technological details.

More than 5% of all accidents involving injury to people take place as the result of a lane change. Therefore, it is sensible to provide the driver with a lane change assistant in order to provide support in this driving maneuver.ISO standard 17387 “Lane Change Decision Aid System” differentiates between three different types of system: The “Blind Spot Warning” monitors the blind spot on the left and right adjacent to the driver’s own vehicle. The “Closing Vehicle Warning” monitors the adjacent lanes to the left and right behind the driver’s own vehicle in order to detect vehicles approaching from behind. The “Lane Change Warning” combines the functions of “Blind Spot Warning” and “Closing Vehicle Warning.”Almost all major vehicle manufacturers are now offering systems that assist the driver to change lanes. Systems with “Blind Spot Warning” are available from Ford, Jaguar, Mercedes-Benz, Nissan/Infiniti, Peugeot, and Volvo. Systems with “Lane Change Warning” are available from Audi, BMW, Mazda, and VW. All vehicle manufacturers use an optical display in or near to the exterior mirrors in order to show information for the driver. The majority of vehicle manufacturers use radar sensors that are installed in the rear of the vehicle. Two-level, escalating driver information is only offered in some of the systems. The type of escalation (optical, acoustic, tactile, lateral guidance intervention) usually differs from one vehicle manufacturer to another.The performance capability of the lane change assistants described above is already quite considerable. However, all of these systems have their limits, and the vehicle manufacturers need to inform drivers of these in the owner’s manual, for example.

Steering and evasion assistance defines a new and future class of driver assistance systems to avoid an impending collision with other traffic participants. Dynamic and kinematic considerations reveal that an evasive steering maneuver has high potential for collision avoidance in many driving situations. Three different system layouts are described: driver-initiated evasion, corrective evasion, and automatic evasion assistance. Since an automatic steering intervention is a challenging and responsible task, the technological requirements for situation analysis and environment perception are stated. Many technical solutions for a steering intervention are conceivable; therefore several actuator concepts are discussed and assessed with respect to human machine interface (HMI) impacts. A short survey of research activities of industry and academia is given. As an example for a research level prototype, the Daimler automatic evasion assistance system for pedestrian protection is presented in detail. Based on binocular stereo vision, crossing pedestrians are detected by fusion of a pedestrian classification module with a 6D-Vision moving object detection module. Time-To-X criticality measures are used for situation analysis and prediction as well as for maneuver decision. Tested on a proving ground, the prototype system is able to decide within a fraction of a second whether to perform automatic braking or evasive steering, at vehicle speeds of urban traffic environment. By this it is shown that automatic steering and evasion assistance comes to reality and will be introduced stepwise to the market.

Pedestrian accidents are an important aspect of vehicle safety in Europe and throughout the world. Therefore, various countries have already passed statutory regulations on pedestrian protection for vehicles. These mainly focus on assessing passive protection measures. Furthermore the installation of a brake assist system as an active safety system is prescribed in European pedestrian protection legislation. This is because of the significant benefit by reducing the collision speed in pedestrian accidents which was proven in studies with real-world accident data. In future, further active measures will significantly contribute to protect pedestrians because with these technologies it is possible to avoid collisions or mitigate their severity. Combining the active and passive safety technologies to an integrated safety approach will be the most important development objective for further reductions of accidents and casualties in the next years. This chapter provides an overview of active pedestrian protection systems and the new challenges faced when developing these systems. At the beginning, an international comparison of pedestrian accidents and the results of analyzing an in-depth accident database are presented. In the next step, current pedestrian protection measures in the field of infrastructure and passive safety are described. The active safety systems mainly consist of environment sensor systems, functional algorithms, and actuating elements. For each of these components selected, realizations will be shown and discussed for their employment in active pedestrian safety systems. A whole system functionality is created by combining these single modules. In this context, two main system strategies have to be distinguished: on the one hand, system strategies that autonomously engage into the driving situation and on the other hand system strategies that draw the driver’s attention to a dangerous situation by presenting a warning. In addition to the development tools new methods for the system test and the benefit calculation have to be engineered. These new tools and test setups will also be presented.

On virtually all motor vehicles, the bodies have been designed and developed in such a way as to achieve the lowest possible drag coefficient values in order to reduce fuel consumption. This trend has resulted in a gentle wedge shape which greatly restricts the driver’s view when maneuvering. Obstacles can only be poorly discerned – if at all.To overcome these problems, ultrasonic-based parking aids were introduced in the European market in the early 1990s. These systems monitor the rear and the front of the vehicle, and warn the driver if there is an obstacle which can cause a collision. Recently, new functions like semiautomatic Parking Assistance have been realized based on the same sensor technology. Such a system automatically steers the vehicle into the parking space while the driver controls the longitudinal movement of his car.In parallel, cameras to monitor the rear of the vehicle have first been introduced on the Japanese market together with central information systems allowing presenting its picture in the center console area. Due to the availability of powerful image processing units, recently multi-camera systems have been launched. These systems fuse the data of four cameras, for example, to create a 360° top-view picture showing the surroundings of the vehicle.Further sensor improvement and system development of both ultrasonic and camera technology as well as sensor data fusion of different technologies will allow new parking and maneuvering functions with increasing automation grade.The chapter starts with basics of ultrasonic and camera technology. Furthermore, it emphasizes on driver assistance systems for parking and for slow maneuvers based on these sensor technologies.

From their beginnings in the car as a tool for A-to-B navigation, digital maps are experiencing an evolution process that will see them at the forefront of new applications designed to improve active safety and manage fuel consumption. These maps will, in effect, become a new vehicle sensor, with a range exceeding that of camera and radar systems, and an ability to work in all weathers and at night. These new maps will need to be more accurate than those used for navigation, and be fused with a minimized set of map attributes to create new vehicle-interpreted precision maps. This chapter will look at the applications that would benefit from these new maps, which in terms of both safety and energy management applications, provide precise knowledge of the road ahead. This allows the vehicles and drivers to be informed of potentially dangerous situations, and take actions based on exact knowledge of future slopes and curves in the road. In energy management terms, the knowledge of road slope will allow the most fuel-efficient routes to be chosen, and can be used to determine the range of Electric and Hybrid Electric Vehicles (EV/HEV), as well as optimizing engine and transmission for fuel efficiency. We will consider how such maps can be created using a number of different technologies, and how this collection methodology impacts their characteristics. As maps evolve and become more “connected,” the possibilities to update them, and access further geographic data and services, will further increase their usefulness.

Every year, thousands of vehicles are involved in crashes which are attributed to the onset of driver drowsiness. As a result, there are numerous drowsy driver assistance systems (DDAS) available on the market; however, many of these technologies rely on a single predictor of driver drowsiness (e.g., eye closures, lane position, steering). Relying on only one predictor of drowsiness makes the system susceptible to periodic intervals in which data is unavailable due to failure of the single sensor, usage outside of the sensor’s envelope of operation, or driver’s individual differences. Driver drowsiness measures can be classified as either driver-based (those measures derived from the human) or vehicle-based (those measures derived from the vehicle). For driver-based measures, PERCLOS (a measure of slow eye closure) is considered to be a robust measure of driver drowsiness. Machine-vision (MV) slow eye-closure sensors have been developed to estimate the percent of eye closure and calculate the PERCLOS value. However, these MV slow eye-closure sensors’ ability to detect the eye closures is challenged by eyewear, ambient illumination, and head movement. For vehicle-based drowsiness metrics, lane position appears to be a key indicator of driver drowsiness. Lane position is typically estimated through MV technology detecting lane edge markings on the forward roadway scene. The absence of lane edge markings on roadways or instances of low contrast between lane markings and the surrounding scene make it difficult for this MV lane position sensing technology to accurately measure the vehicle’s position within the lane. Typical causes of low contrast lane markings include poor lane marking quality, artificial overhead lighting, or headlight “blooming.” Therefore, a multi-measure approach, that uses multiple distinct sensors, can offer not only sensor redundancy but also provide a data fusion approach whereby both measures provide more robust drowsiness detection than either measure could alone. This chapter describes the salient measures of driver drowsiness, the concept of data fusion in DDASs, and provides a case study of a prototype DDAS that integrates two drowsiness metrics (i.e., PERCLOS and Lane Position) to form an enhanced drowsiness estimate that may prove to be a more robust measure in a real-world, field application as compared to a single metric system.

This chapter presents a real-time computer vision system for monitoring drowsy driver. It uses one remotely located charge coupled device (CCD) camera to acquire video of the driver’s face. From the video, various computer vision algorithms are employed to simultaneously, nonintrusively, and in real time recognize the facial behaviors that closely relate to the driver’s level of vigilance. The facial behaviors include rigid head movement (characterized by 3D face pose), nonrigid facial muscular movement (characterized by facial expressions), and eye gaze movement. The system was tested in a simulating environment with different subjects and it was found robust, reliable, and accurate in characterizing facial behaviors.

Drowsy and fatigue driving is a major transportation safety concern and is responsible for thousands of accidents and numerous fatalities every year. The resulting harms of drowsy/fatigue driving could be even higher among commercial vehicles. Drowsy driving crashes are usually of high severity due to the drivers’ significant loss of control, often leading to unpredicted vehicle trajectory and no braking response. Reliable safety systems are needed to mitigate these crashes. The most important challenge is to detect the driver’s condition sufficiently early, prior to the onset of sleep, to avoid collisions.Various detection methods have been proposed by researchers and a few systems are available in the commercial market. In general, drowsiness detection methods fall into two major categories of monitoring physiological and physical conditions of the drivers and monitoring vehicle-related variables based on driver control functions that correlate with the driver’s level of drowsiness. Each method has its advantages and shortcomings. A reliable detection method needs to be integrated with a safety system which may include advisory warning, semi-control, or full control of vehicle, i.e., braking and steering to achieve safe conditions. The type and intensity of warning or control should also be carefully selected and are discussed in another chapter.This chapter first reviews the statistical significance of the crash data due to drowsiness and fatigue conditions. Then, the issues concerning various detection methods are discussed. Detection systems based on driver control functions are mainly discussed in this chapter. The concepts and approaches presented in this section are from a comprehensive literature review including the author’s past research; they can guide the development of safety systems for a passenger or commercial vehicles.

Driving under the influence of fatigue and sleepiness is a serious safety concern. Hundreds of lives and billions of dollars are lost every year due to accidents caused by driver drowsiness. There are many aspects to the problem of a driver falling asleep while driving that include causes, detection, monitoring, warning, and countermeasures against drowsy driving. A number of crucial design issues have to be considered before the anticipated benefits of the drowsy driving warning can be fully realized. In this chapter the two major aspects, that is, warning and countermeasures, are discussed.Warning means to convey to the driver about his/her state of sleepiness/drowsiness so that corrective actions can be taken. There are many issues related to warning system design but the two main concerns are when and how to warn the driver, that is, alarm modality and alarm timing. Although there are no standard guidelines for selection and design of appropriate alarm modalities, at least three types of modalities (visual, audio, and haptic/tactile) and their combinations are possible for any alarm design. An important component of collision avoidance system is the algorithm that determines the timing of warning. A poorly timed alarm may actually undermine the safety of the driver. An alert issued too early may be ignored by drivers if they are unable to perceive the cause of the warning. On the other hand, if it occurs too late, it may be viewed as ineffective.An alarm that does not represent the true state of driver drowsiness, that is, the driver is not drowsy but the system issues a warning, is called false alarm. False and nuisance alarms are a particular problem for automotive collision avoidance and warning systems.A comprehensive review of the literature regarding driver fatigue/drowsiness warning research, the present state of research and technologies being developed, and issues related to warning/alarm design and the future trends are highlighted. Driver fatigue-related countermeasures are also discussed along with their merits and demerits.

In developing a vision system for a vehicle, different setup constraints and issues must be considered.Space, wiring, or lighting are also typical issues to be also faced in industrial scenarios; nevertheless, when a vision system has to be deployed inside a vehicle they have to be more carefully studied and often drive the hardware selection.Moreover, cameras are to be installed on moving vehicles and this led to additional problems to be faced. In fact, camera movements, oscillations and vibrations, or different and even extreme illumination conditions have to be taken in account when developing machine vision software.

This chapter starts with a theoretical definition of the sensor technology and describes the main parameters suitable to classify cameras. The section continues with a brief discussion on optics and sensors. A section then is dedicated to camera-specific software such as firmwares, API, or dedicated SDK. The last section describes mechanical issues specific to vehicular applications.

The localization of painted road markings is a key aspect of environment reconstruction in urban, rural, and highway areas, allowing a precise definition of the safely drivable area in front of the vehicle.Lane detection algorithms are largely exploited by active safety systems in the automotive field, with the aim of warning the driver against unintended road departures, but are also essential in fully autonomous vehicles, since they complement the data coming from other sources, like digital maps, making it possible to navigate precisely even in complex scenarios.This chapter introduces potential approaches and requirements for lane detection and describes in detail one of such algorithms and its results.

Obstacle detection is a widely studied field in the automotive industry because of the great importance it assumes in all systems that provide autonomous navigation of vehicles in an environment.Many different obstacle detection systems have been developed. The main differences between these systems are the types of algorithms and sensors employed. Many studies have focused on road obstacle detection in order to perform such tasks as pre-crash, collision mitigation, stop and go, obstacle avoidance, and inter-distance management. An important issue in ensuring the reliability of obstacle detection is the choice of sensors: digital cameras, infrared sensors, laser scanners, radar, and sonar are commonly used to provide a complete representation of the vehicle’s surrounding area, allowing interaction with the world. Inertial sensors like odometers, speed sensors, position sensors, accelerometers, and tilt sensors are used to monitor the motion of the vehicle, measuring its speed, orientation, and position. This chapter is structured in three main sections: The first will introduce a classification of all perception sensors that can be used in this field; a brief description of each sensor will be provided in order to underline pros and cons regarding the obstacle detection field. In the second section, the main algorithms of obstacle detection will be shown, classified by the kind of sensor (or sensors) employed. The third section presents obstacle detection systems that use sensory fusion combining artificial vision with distance detection sensors like laser or radar.

The system described in this chapter is a traffic sign recognition based on a color camera. Each algorithm step will be detailed: a color segmentation to identify the possible regions of interest, a shape detection, and the final sign classification and tracking. A description of the encountered problems and their solutions is given as well. The last section presents the algorithm results.

In this chapter, a description of ACC system will be given, focusing on how to develop this kind of system using optical sensors and computer vision techniques. In particular, two approaches based on different sensors fusion (LIDAR, radar, and camera) are described.

These sections introduce a vision-based system designed for monitoring the area that a driver cannot see from exterior mirrors, usually referred to as blind spot.This is a challenging task that requires to discriminate from vehicles and background when both are not static and also to cope with usual automotive problems like camera vibrations and oscillations.

Outlines the primary requirements and constraints applicable to vehicular communications. Describes key operational issues and summarizes typical applications. Provides overview of Dedicated Short range communications (DSRC) and describes typical implementation configurations.

Cooperative vehicle to infrastructure and vehicle to vehicle safety and mobility applications are described. Applications are separated into static and dynamic categories, such as fixed roadway hazards versus changing traffic signals. Communications requirements are discussed in the context of overall application functional requirements including the specific safe distances required for driver to react to alerts and warnings and to take appropriate action. These requirements are translated into conventional link budget terms relative to those that affect communication reliability such as a transmitter power and receiver sensitivity, multipath fading, and overall channel capacity. The implications of networked transmissions versus broadcast, non-network transmissions are discussed in the context of safety applications.

Positioning and communications technologies are enabling the collection of massive amounts of probe data from the vehicle fleet. The quality and quantity of the data vary significantly depending on technology and collection interval, with the best data coming from GPS systems integrated into vehicles. Real-time data collection provides, for the first time, detailed insight into vehicle behavior throughout the major elements of the transportation network. This is used by roadway managers to optimize performance of the infrastructure and by intelligent vehicles to increase situation awareness beyond the range of their autonomous sensors, potentially leading to significant increases in safety and efficiency.Historical probe data can be used to create a map of driver behavior at every point in the transportation system. Behavioral mapBehavioral maps share some attributes with traditional physical maps, but other attributes, such as average vehicle speed and distribution of speeds, enable novel implementations of intelligent vehicle applications. Behavioral data can be used directly to define normal or acceptable behaviors. In addition, a particular driver’s preferred location on behavior distribution curves can be used to predict future behavior and personalize interactions.Probe data will become more prevalent and valuable as penetration increases and latencies decrease. Ultimately, the need for probe data will help motivate deployment of short-range, high data rate communications between vehicles and with infrastructure.

Security is an essential part of all vehicle networks. Communication among vehicles and roadside infrastructure needs to be secure, preserve vehicle privacy, and support efficient and effective removal of bad actors. The threat model for vehicle networks describes three categories of threat agents whose motives range from obtaining preferential treatment to tracking vehicles and disrupting transportation. Vehicle and roadside equipment, wireless communications, and network and software technologies are vulnerable to attack. The notion of privacy in vehicle networks encompasses the properties of anonymity and unlinkability. Vehicle tracking is a privacy threat that exploits vehicle communications, application transactions, and roadway conditions. Public Key Infrastructure is the predominant security architecture among vehicle networks, providing message authentication, integrity protection, and data encryption. The certificate management scheme affects privacy, the removal of bad actors, and system robustness. The combinatorial certificate scheme used in the US DOT proof-of-concept trial is an example of a shared certificate scheme. Removing bad actors in shared certificate schemes is challenging. Certificate revocation may affect many innocent vehicles, which may lose their network privileges. The short-lived, unlinked certificate scheme is an example of a unique certificate scheme that avoids the “one affects many” problem. It separates the certificate authority authorization and assignment functions and issues a large number of short-lived certificates, where certificate expiration may eliminate the need for revocation. Efficient and effective intrusion detection is critical to maintaining vehicle network integrity. Vehicle and roadside equipment, the certificate authority, application servers, and other network-based systems can participate in intrusion detection.

An ITS system will feature many different types of application. These applications all have their own security and performance needs, which may differ from each other. Additionally, the fact that different applications may coexist on the same device introduces additional security considerations. This chapter reviews the security mechanisms that may be used for different classes of application, and for the device as a whole, and surveys their deployment history and their support in standards. The aim is to provide an implementer of any ITS application with a usable starting point to help them determine which security services to use in their application and how those services should be implemented. Particular attention is paid to issues of privacy: ITS applications have an inherent risk of revealing personal data, such as current location, to parties who have no right to that data, and as such an implementer must take care to ensure that privacy is preserved to at least the level required by local regulations. The chapter also reviews security management operations such as issuing and revoking digital certificates.

Vehicles are evolving into autonomous mobile-connected platforms. The rationale resides on the political and economic will towards a sustainable environment as well as advances in information and communication technologies that are rapidly being introduced into modern passenger vehicles. From a user perspective, safety and convenience are always a major concern. Further, new vehicles should enable people to drive that presently can not as well as to facilitate the continued mobility of the aging population.Advances are led by endeavors from vehicle manufacturers, the military and academia and development of sensors applicable to ground vehicles. Initially, the motivators are detailed on the reasons that vehicles are being built with intelligent capabilities. An outline of the navigation problem is presented to provide an understanding of the functions needed for a vehicle to navigate autonomously. In order to provide an overall perspective on how technology is converging towards vehicles with autonomous capabilities, advances have been classified into driver centric, network centric and vehicle centric. Vehicle manufacturers are introducing at a rapid pace Advanced Driving Assistance Systems; these are considered as Driver Centric with all functions facilitating driver awareness. This has resulted on the introduction of perception sensors utilizable in traffic situations and technologies that are advancing from simple (targeted to inform drivers) towards the control of the vehicle. The introduction of wireless links onboard vehicles should enable the sharing of information and thus enlarge the situational awareness of drivers as the perceived area is enlarged. Network Centric vehicles provide the means to perceive areas that vehicle onboard sensors alone can not observe and thus grant functions that allow for the deployment of vehicles with autonomous capabilities. Finally, vehicle centric functions are examined; these apply directly to the deployment of autonomous vehicles. Efforts in this realm are not new and thus fundamental work in this area is included. Sensors capable to detect objects in the road network are identified as dictating the pace of developments.The availability of intelligent sensors, advanced digital maps, and wireless communications technologies together with the availability of electric vehicles should allow for deployment on public streets without any environment modification. Likely, there will first be self-driving cars followed by environment modifications to facilitate their deployment.

In order to safely navigate in a dynamic environment, a robot requires to know how the objects populating it will move in the future. Since this knowledge is seldom available, it is necessary to resort to motion prediction algorithms. Due to the difficulty of modeling the various factors that determine motion (e.g., internal state, perception), this is often done by applying machine-learning techniques to build a statistical model, using as input a collection of trajectories array through a sensor (e.g., camera, laser scanner), and then using that model to predict further motion.This section describes the basic concepts involved in current motion learning and prediction approaches. After introducing the Bayes filter, it discusses Growing Hidden Markov Models, an approach which is able to perform lifelong learning, continuously updating its knowledge as more data are available. In experimental evaluation against two other state-of-the-art approaches, the presented approach consistently outperforms them regarding both prediction accuracy and model parsimony.The section concludes with an overview of the current challenges and future research directions for motion modeling and learning algorithms.

This chapter describes an algorithm for determining the speed and the attitude of a sensor assembling constituted by a monocular camera and inertial sensors (three orthogonal accelerometers and three orthogonal gyroscopes). The system moves in a 3D unknown environment. The algorithm inputs are the visual and inertial measurements during a very short time interval. The outputs are the speed and attitude, the absolute scale and the bias affecting the inertial measurements. The determination of these outputs is obtained by a simple closed-form solution which analytically expresses the previous physical quantities in terms of the sensor measurements. This closed-form determination allows performing the overall estimation in a very short time interval and without the need of any initialization or prior knowledge. This is a key advantage since allows eliminating the drift on the absolute scale and on the orientation. The performance of the proposed algorithm is evaluated with real experiments.

Autonomous navigation using a single camera is a challenging and active field of research. Among the different approaches, visual memory-based navigation strategies have gained increasing interests in the last few years. They consist of representing the mobile robot environment with visual features topologically organized gathered in a database (visual memory). Basically, the navigation process from a visual memory can be split in three stages: (1) visual memory acquisition, (2) initial localization, and (3) path planning and following (refer to Fig. 53.1). Importantly, this frame work allows accurate autonomous navigation without using explicitly a loop closure strategy. The goal of this chapter is to provide to the reader an illustrative example of such a strategy.

This chapter describes detection and tracking of moving objects (DATMO) for purposes of autonomous driving. DATMO provides awareness of road scene participants, which is important in order to make safe driving decisions and abide by the rules of the road. Three main classes of DATMO approaches are identified and discussed. First is the traditional approach, which includes data segmentation, data association, and filtering using primarily Kalman filters. Recent work within this class of approaches has focused on pattern recognition techniques. The second class is the model-based approach, which performs inference directly on the sensor data without segmentation and association steps. This approach utilizes geometric object models and relies on non-parametric filters for inference. Finally, the third class is the grid-based approach, which starts by constructing a low level grid representation of the dynamic environment. The resulting representation is immediately useful for determining free navigable space within the dynamic environment. Grid construction can be followed by segmentation, association, and filtering steps to provide object level representation of the scene. The chapter introduces main concepts, reviews relevant sensor technologies, and provides extensive references to recent work in the field. The chapter also provides a taxonomy of DATMO applications based on road scene environment and outlines requirements for each application.

This chapter addresses safe mobile robot navigation in complex environments. The challenges in this class of navigation problems include nontrivial vehicle dynamics and terrain interaction, static and dynamic environments, and incomplete information.This complexity prompted the design of hierarchical solutions featuring a multilevel strategy where strategic behaviors are planned at a global scale and tactical or safety decisions are made at a local scale. While the task of the high level is generally to compute the sequence of waypoints or waystates to reach the goal, the local planner computes the actual trajectory that will be executed by the system. Due to computational resource limitations, finite sensing horizon, and temporal constraints of mobile robots, the local trajectory is only partially computed to provide a motion that makes progress toward the goal state. This chapter focuses on safely and efficiently computing the local trajectory in the context of mobile robot navigation.This chapter is divided into three sections: motion safety, iterative motion planning, and applications. Motion safety discusses the issues related to determining if a trajectory is safely traversable by a mobile robot. Iterative motion planning reviews developments in local motion planning search space design with a focus on potential field, sampling, and graph search techniques. The applications section surveys experiments and applications in autonomous mobile robot navigation in outdoor and urban environments.

This chapter addresses autonomous navigation in populated and dynamic environments. Unlike static or controlled environments where global path planning approaches are suitable, dealing with highly dynamic and uncertain environments requires to address simultaneously many difficult issues: the detection and tracking of the moving obstacles, the prediction of the future state of the world, and the online motion planning and navigation. In the last few years, the problem of incomplete, uncertain, and changing information in the navigation problem domain has gained even more interest in the robotic community and probabilistic frameworks aiming to integrate and elaborate properly such information have been developed. This chapter is divided into three sections: First section introduces the main challenge of this approach. Section 2 focuses on navigation using prediction of the near future and Sect. 3 discusses on integrating human in the navigation decision scheme.

The development of autonomous vehicles garnered an increasing amount of attention in recent years. The interest for automotive industries is to produce safer and more user-friendly cars. A common reason behind most traffic accidents is the failure on the part of the driver to adequately monitor the vehicle’s surroundings. This chapter addresses the problem of estimating the collision risk for a vehicle for the next few seconds in urban traffic conditions.Current commercially available crash warning systems are usually equipped with radar-based sensors on the front, rear, or sides to measure the velocity and distance to obstacles. The algorithms for determining the risk of collision are based on variants of time-to-collision (TTC). However, it might be misleading in situations where the roads are curved and the assumption that motion is linear does not hold. In these situations, the risk tends to be underestimated. Furthermore, instances of roads which are not straight can be commonly found in urban environments, like the roundabout or cross-junctions.An argument of this chapter is that simply knowing that there is an object at a certain location at a specific instance in time does not provide sufficient information to assess its safety. A framework for understanding behaviors of vehicle motion is indispensable. In addition, environmental constraints should be taken into account especially for urban traffic environments.This chapter proposes a complete probabilistic model motion at the trajectory level based on the Gaussian Process (GP). Its advantage over current methods is that it is able to express future motion independently of state space discretization. Driving behaviors are modeled with a variant of the Hidden Markov Model. The combination of these two models provides a complete probabilistic model for vehicle evolution in time. Additionally a general method of probabilistically evaluating collision risk is presented, where different forms of risk values with different semantics can be obtained, depending on its applications.

While legal issues of driver assistance systems appear to be largely solved for systems supporting the driver with information or such that remain easily overrideable/oversteerable, the increase in automation can eventually bring about a paradigmatic change of the “driving task”: Up to date, the driver’s responsibility for the use of systems is maintained, thus remaining within the traditional concept of driving. In future, however, a substantial further increase in automation can lead to a structural shift. The legal issues this change would raise must be identified and handled at an early stage of research to avoid false investment as well as inequitable legal consequences. The legal issues related to driver assistance and autonomous systems are thereby cross-sectional in nature and have a link to the issue of acceptance as far as a basic legal change is intended and necessary.Cooperative systems, presently under intensive research with their plentiful possibilities and benefits, also give rise to legal uncertainties in several fields (including liability). However, here communication architecture, technical design, as well as potential operators vary tremendously depending on use case and information required for the respective function. This retards a universally valid description of accompanying legal issues. As long as these systems are, however, only meant to take effect by informing the driver on the oncoming traffic situation and dangers without suggesting full reliability (as is presently mostly under research), data privacy should prove to be the only important (but resolvable) issue.

In this chapter, we will examine how the development of ITS can help solve major problems of humanity which are energy, climate change, congestion, and safety.The coming decades will accompany a mobility revolution and will see the emergence of a new relationship of citizens with their automobile.The concept of sustainable mobility covers deployment of electro-mobility but also new forms of relationships between citizens and nature.Information and communication technologies will help promote multimodality, ticketing, car sharing, carpooling, PRTs, and progressive automation of driving through the dissemination of Advanced Driver Assistances Systems (ADAS) and cooperative systems (V2V and V2I).

The technologies for implementing intelligence in road vehicles are improving very rapidly and products are now available in the market. Not only for traditional passenger vehicles but also for trucks and buses and for a new generation of urban vehicles that are now appearing on the market as a transportation service instead of a product to be bought by the final user.This chapter will look into the potential markets for the technologies in these different applications such as:Private passenger vehiclesPublic urban vehiclesUrban freight deliveryRapid transit (including BRT, PRT, and CTS)Long-distance freight