Automated Driving

This chapter introduces the emerging topic of automated driving and motivates the subsequent chapters written by renowned authors and experts from industry and academia. We briefly discuss essential building blocks and key technologies to enable automated driving. Furthermore, we address required advances, inspirations, and fields of actions to shift automation to the next level of maturity and hence to foster end-user acceptance of this exciting and disruptive technology.

The goal of this chapter is to give a high-level overview about the importance of privacy and security provisions in autonomous vehicles. It describes where we currently are, where the challenges with security and privacy may be in the context of autonomous vehicles, and what should be done to deal with those challenges. Thereby, it is also important to take the user into account, as he/she has to trust the autonomous vehicle and the company behind it; otherwise autonomous vehicles will not get accepted. Trust here is related to security, privacy but also safety

This chapter provides a short overview of driving automation for on-road vehicles from the standardization point of view with a focus on various levels of automation. The content of this chapter gives information about current technical standards and future standardization topics. Furthermore, it gives a plea for worldwide harmonized standardization and legislative efforts regarding automated driving in near future. The main target is to reach a harmonized regulatory approach for automated driving at the European Union (EU) level that would allow for innovation and, also, improve safety, efficiency, and environment while preserving the comfort of the passengers.

This survey focuses on trajectory tracking controllers of advanced driver assistance functions for comfortable and safe automated driving on highways. A short introduction to today’s driver assistance functions and their control objectives is given. Different control schemes that have been proposed during the past few years are discussed, including essentially PID control, fuzzy control, optimal state feedback controllers, sliding mode control, and model predictive control. The separation of longitudinal and lateral dynamics as well as their combination is tackled. In addition to the control design for assistance functions, this work lists prominent controllers of autonomous vehicle prototypes. A simulation of a highway scenario compares the performance of the different control approaches.

In the United States, the DARPA (Defense Advanced Research Projects Agency) Grand Challenges [Thrun et al. (J. Field Robot. 23(9):661–692, 2006); Urmson et al. (J. Field Robot. 25(8):425–466, 2008); Urmson et al. (J. Field Robot. 23(8):467–508, 2006); Campbell (Steering Control of an Autonomous Ground Vehicle with Application to the DARPA Urban Challenge. Massachusetts Institute of Technology, 2007)] demonstrated that autonomous driving can be achieved through vision and sensor systems capable of detecting and interpreting the vehicle operating environment, rather than through autonomous driving options relying on the infrastructure (e.g., through magnets installed on the road surface to indicate the lanes), or vehicle-to-vehicle or vehicle-to-infrastructure communication systems. The latter options are very useful to further enhance the performance, safety, and energy efficiency, but are not strictly required. In a typical automated driving system, a reference path and a reference speed profile are defined based on the sensed environment. At a lower level of the control system hierarchy, a path tracking controller is responsible for calculating the steering angle for achieving the reference trajectory, while a speed controller determines the wheel torque demand for tracking the reference speed. Speed control implementations are already quite common in production vehicles equipped with cruise control and adaptive cruise control systems. Hence, the core element of novelty for autonomous driving in the area of vehicle control is represented by the steering control function for path tracking. Different steering-based path tracking algorithms, ranging from geometrical methods to model-predictive controllers, are presented and discussed in this contribution, together with the expected future research and vehicle implementation directions in the field.

The presented paper discusses a concept of a model-based reference lane calculation method for autonomous vehicle guidance. Recently shown reference lane calculation methods for autonomous passenger cars without trailer attached were generally performed by a simple “mid-lane” guidance approach, which is sufficient for these types of vehicles to keep them in track. For vehicles with more complex driving dynamics like heavy truck–semitrailer combinations or vehicles with attached trailers, however, a more sophisticated approach for reference lane calculation is necessary to achieve all requirements regarding driving safety and driving comfort for high-quality driving assistance. The presented paper therefore focuses on the task of designing a reference lane calculation method for vehicles with complex driving dynamics, which leads to a more general approach for determining the reference lane to achieve all the requirements of safe and comfortable lane-keeping tasks.

In order to facilitate automated driving, a reliable representation of a vehicle’s environment is required. This chapter provides a survey of techniques for the perception of both static and dynamic environments including key algorithms for object tracking and data fusion. In addition, the particular challenges of this field from a practitioner’s perspective are discussed and compared to the state-of-the-art design and implementation paradigms.

This chapter presents a test infrastructure for the Galileo satellite system and its applications in advanced driver assistance systems. For this purpose, we first introduce a sensor data fusion of Galileo signals and vehicle data using an extended Kalman filter. The fused data give an estimation of vehicle states and position. Based on the introduced sensor fusion, we show two applications including results from experiments we conducted in the aforementioned test infrastructure. The first application is a cooperative adaptive cruise control system, which uses navigation data in combination with digital road maps as well as V2V communication. The second application is a collision avoidance system, which uses both navigation data and inertial data to estimate the relevant vehicle states for a controller to let the vehicle follow a given evasion path.

Modern passenger vehicles increasingly incorporate perception systems that allow the deployment of advanced driving assistance systems which in turn shall led to highly automated systems and ultimately to autonomous vehicles. Despite the numerous advances in sensors and communications technologies applied to passenger vehicles, perception remains a major challenge due to the complexity of the task, the vehicle geometric constraints and cost. Digital navigation maps have proven themselves to be essential for driver guidance and have replaced paper maps. They store the geometric description of roads and associated features. Due to the limits of perception systems, errors occur on the understanding and relevance of the detected objects, when using multiple sensors discordances occur that lead to unknowns. By projecting sensor information from the perceived environment into digital navigation maps, information can be contextualized. The resulting representation of the world is much easier to interpret by a machine, but also to ensure the coherence of the perceived information.This article deals with two aspects of perception: situation understanding and the detection of errors in maps. One goal is to demonstrate how contextualized information can be used to facilitate situation understanding, that is to provide meaning to the spatio-temporal relationship between perceived objects, road features, and the subject vehicle. Situation understanding is a fundamental feature for a machine to take decisions in an appropriate manner, particularly, if the navigation of the vehicle is to be governed by it. The approach is based on the application of ontologies to facilitate contextual descriptions by enabling the formalization of information in a semantic manner. The digital map on board the vehicle stores a priori knowledge about the environment, that is the underlying structure and static context. Digital maps contain different errors, particularly regarding road geometry. In case of errors, the association between the perceived objects and maps for situation understanding will be wrong. This will propagate through the system and lead to hazardous situations. The chapter is thus completed by introducing a mathematical formalism that allows for the detection of geometric map errors in real time. The uniqueness of this formalism is that it uses production type components, namely GNSS receivers and vehicle state data. The approach includes a mechanism that allows for the identification of the error, which could be in the digital map itself or due to errors in the position estimates of the vehicle.The chapter combines theoretical developments with experimental data, as the proposed solutions were tested in public road networks using purposely equipped vehicles to demonstrate the viability and advantages of the theoretical developments.

Automotive radar based on millimetre waves—today in the 24/26 GHz and in the 77/79 GHz range—has been under investigation and development since several decades. Already in the early 1970s first 35 GHz radar sensors were tested over several millions of road kilometres. Since 1998—beginning with an ACC (Autonomous Cruise Control) radar sensor in the Mercedes-Benz S-class sedan—automotive radar is commercially available and employed by various OEM’s all over the world. These days such radar systems are used for various vehicular applications, predominantly for functions like ACC, or BSD (Blind Spot Detection) to name just two important safety functions. The development over the last four decades is described.

Driver assistance systems have been successfully deployed to the market in the last 15 years, resulting in an increase of driving comfort and driving safety. In the future, these systems will be able to analyze ever more complex traffic situations and to support the driver or even act independently. Upcoming functionality will combine longitudinal and lateral control to partially automated driving functions; highly automated functions will soon follow. With the increase of automation, the role of the driver is going to gradually change from an active driver to a passenger, at least for some duration of the drive. In this chapter, we discuss the implications of this evolution on the requirements for future vehicle architectures.

Automated driving (A.D.) requires concurrent execution of multiple complex driving functions on automotive embedded platforms. In general, such systems can be partitioned into early stages including sensor processing, individual perception, and cognition functions and into later, more centralized stages that perform data fusion, planning, and decision making. In this chapter, we exemplarily concentrate on automotive embedded processing systems for perception and cognition problems, however, we expect similar problems also on later stages such as data fusion. For perception and cognition, one can observe a wide gap between required processing power and the achievable embedded realizations which have to fulfill non-functional requirements such as low power and small cost. Furthermore, these systems must perform all processing under strict safety requirements that guarantee deadlines and provide high system robustness.

This chapter provides practical insights into specific systems engineering and architecture considerations for building autonomous driving systems. It is aimed at the ambitious practitioner with a solid engineering background. We envision such a practitioner to be interested not just in concrete system implementations, but also in borrowing ideas from the general theory of intelligent systems to advance the state of autonomous driving.

This paper discusses the need and requirements for an open dependable power-computing (DPC) platform (including operating system, middleware, update process, etc.) to support advanced assisted and automated driving functions. Automated driving functions pose a set of new requirements for an in-vehicle computing platform (e.g., higher computational efficiency, high resource demands, etc.).We believe that one aspect of coping with these new requirements is openness and therefore argue for establishing an open, common, and extensible high-quality computing platform.Here we discuss what we mean by openness, what we see as the main requirements for such a platform, and how such a platform could be realized.

Highly Automated Driving (HAD) opens up new middle-term perspectives in mobility and is, therefore, currently one of the main goal in the development of future vehicles. In particular, premium manufacturers, such as the BMW Group, put Highly Automated Driving at the top of the roadmap. This chapter details the motivation behind Highly Automated Driving from a road safety perspective. Assessing the effect of HAD functions on road safety is essential for the homologation of such complex systems. New methods are needed to enable the assessment of complex driving functions and demonstrate the increase in road safety. This problem will be considered and to a possible approach will be referred. Furthermore, the additional (indirect) safety benefit will be described through the usage of HAD technology to improve Active Safety Systems.

Today’s innovative automated driving systems (ADS) functions are realised by highly interconnected and networking cyber-physical systems based on existing automated driving assistance systems (ADAS). These interconnections increase the complexity of so-called systems of systems, because automation requires information and interaction with its environment. All possible interactions must be known for the definition of the intended system behaviour in order to identify any malfunctions of ADS, which may propagate over the system boundaries and influence other systems to fail in a harmful way. Hidden links are able to affect unwanted operational system states so that they cannot be perceived as failure modes. For that reason, functional safety is an important topic for reduction of safety-critical risk to cause failures in complex automotive systems.The chapter presented discusses the application of the automotive functional safety standard ISO 26262 in context of ADS. The following main topics are highlighted: Complexity of automated driving systems, issues concerning availability and reliability, importance of the concept phase and the role of the driver. Furthermore, proposals are made on how to handle these challenges and for feasible enhancements of the current ISO 26262 standard. Existing and promising methods are discussed that deal with the increasing complexity for the development of future ADS.

There is a difference between developing a vehicle that is driving itself as safely as today’s cars are driven and assessing this vehicle in terms of safety. The current approach to prove this safety is mainly and ultimately based on road testing.This chapter first explains why proving safety of automated vehicles for their first introduction will be economically challenging when using road tests. The argumentation is based on a theory using a statistical approach. In the second part of the chapter, further conclusions are drawn from these findings, and a new universal theory on the introduction of automated systems motivated by an extended statistical point of view is proposed.

Due to the constantly growing global population and the ageing society, the CO₂ production as well as the demand for mobility is constantly rising and hence needs new concepts as offered by advanced driver assistance systems up to fully automated vehicles. If existing approaches were used for these new concepts, the validation of these systems would require unacceptable long validation times and high costs. Therefore, new validation methods are necessary. The extensive use of validation in mixed real and virtual environments together with statistical methods of combinatorial testing offers promising solutions. Standardization of scenarios and learning validation schemes can reduce the validation effort to an acceptable level.

During the last 25 years, the driving abilities of automated vehicles have progressed rapidly. This went along with a huge increase of complexity for automated vehicles, regarding the multiplicity of interacting components being required to implement the functionality of automated vehicles.

Advanced driver assistance and automated driving can influence traffic safety in a variety of ways. The development and implementation of safety-relevant functions require prospective, quantitative assessment of their traffic safety impacts. Both benefits and risks can be quantified using simulation-based virtual experimental techniques. To this end, traffic phenomena are modeled taking into account key safety-relevant processes; “stochastic” simulation is performed on large, representative virtual samples. The virtual representations of traffic phenomena are based on detailed, stochastic models of drivers, vehicles, traffic flow, and the road environment, together with their interactions. The models incorporate knowledge from field operational test (FOT), naturalistic driving studies (NDS), laboratory and driving simulator experiments, and other sources. Simulation serves to synthesize this knowledge. Large-scale, comprehensive simulations could help in identifying and evaluating the relevant situations in which automated driving impacts traffic safety. One key objective is a standardized harmonized methodology, agreed upon by all stakeholders, for comprehensive assessment of the impact of new driver assistance or automated driving functions on traffic safety.

Increasing automation requires a change from looking into the driver performance in single situations to the analysis of the entire likelihood of risks. Driver performance, system performance, and the traffic in general are important issues in this. The approach to develop safety in use, where these components have to be considered, is introduced and explained. The difference to functional safety according to the ISO 26262 is also made explicit. The analysis of safety in use offers the opportunity of quantifying risks. An acceptable reference value for this risk does not yet exist. Two existing approaches that could be used as a reference are explained, dealing with similar risks. In the second part of this paper, methods and examples of how to get necessary data for an analysis of safety in use are described. Literature reviews, questionnaires, studies in driving simulator or in real vehicles, and observation of traffic or field operational tests are explained using examples of conducted investigations in China and Germany.

Proving techniques and methods for safety critical systems in order to ensure a certain behavior as well as their corresponding safety requirements has still been a challenge for many years. Although the current situation in many areas like the automotive industry has improved a lot, new challenges are in sight especially when considering autonomous and adaptive systems approaching. Such systems have to reason about the current state and stimuli from their environment without humans in the loop or are allowed to change their behavior over time. Such systems induce new requirements for quality assurance and in particular testing. Here the focus has to be on providing guarantees of a wanted behavior before deployment of the systems even in case of changes or failures that might arise at runtime. In this paper, we discuss the underlying challenges and potential feasible solutions. In addition, we highlight similarities and differences with the current situation of testing safety critical systems.

The European research project AdaptIVe aims to achieve major breakthroughs leading to more efficient and safer automated driving. The project develops and tests automated driving functions for cars and trucks. The new technologies will be demonstrated in seven passenger cars and one truck being built-up by the project. The research covers several scenarios, including motorways, urban environment, and close-distance manoeuvres. In parallel, the project defines specific evaluation methodologies and addresses legal aspects.

The Drive Me project focuses on studying potential benefits when autonomous vehicles are introduced on larger scale in the road transportation system. It aims to put a fleet of 100 autonomous vehicles in the hands of ordinary Volvo customers to operate on public roads in Gothenburg, Sweden, in 2017. The customers will not need to continuously supervise the vehicle operation and therefore will be allowed to spend time on other activities. The autonomous vehicles will be used as measurement probes for research on the effect on safety, traffic flow, and energy efficiency. Thus, the Drive Me project has a high-profile ambition to define and evaluate how autonomous vehicles will have a major importance for quality of life and achievement of a sustainable urban environment.

The presented paper presents the ongoing Swedish national research project FUSE (FUnctional Safety and Evolvable architectures for autonomy). Some of the research questions addressed in this project are summarized. The research questions are related both to functional safety and the E/E architecture of vehicles aimed for higher degrees of automation, including fully autonomous ones.

Automated driving is expected to significantly contribute to future safe and efficient mobility. Whereas classical automated approaches solely consider the host vehicle, AutoNet2030 aims to investigate a cooperative approach where communication is used to build decentralized control systems, facilitate cooperative traffic flow optimization, and enhance perception. This chapter introduces the concepts and methodology of AutoNet2030 in order to contribute to a cost-optimized and widely deployable automated driving technology.

Machines are converging towards autonomy. The transition is driven by safety, efficiency, environmental and traditional ‘robotics automation concerns’ (dirty, dull and dangerous applications). Similar trends are seen in several domains including heavy vehicles, cars and aircraft. This transition is, however, facing multiple challenges including how to gradually evolve from current architectures to autonomous systems, limitations in legislation and safety standards, test and verification methodology and human–machine interaction.

To address the safety challenges arising from future mobility systems requirements, novel analysis methods and tools are needed. Besides the evolution and utilization of new hardware architectures, software development must address the increasing complexity of new highly automated mobility solutions. Consequently, the single most important roadblock for this market is the ability to come up with an affordable, safe multi-core development methodology that allows industry to deliver trustworthy new functions at competitive prices. The ITEA3 ASSUME project delivers solutions for the development and verification of highly automated, safety relevant, and performance critical mobility systems. The ASSUME consortium includes leading European industry partners for mobility solutions and tool and service providers for embedded system development, as well as leading research institutes for static analysis in model-driven and traditional embedded systems development.

This industrial project introduces an ultraflat automated robot that can be used for testing driver assistance systems as well as automated driving scenarios. Because of its very stable and flat structure, it can be overrun by test vehicles without any damage. Therefore, it is possible to use this robot for both pre- and postcrash testing scenarios as well as for the evaluation of active safety systems (e.g., automatic brake) and autonomous driving.

As urban populations increase, issues such as traffic jams, pollution and road fatalities will grow in tandem. More than half of the world’s population now lives in urban areas (54 %), and growth is expected to accelerate in years to come.

ARTEMIS-IA is THE European association for actors in Embedded Intelligence. The association implements the ARTEMIS European Technology Platform (ETP) activities, which brings together R&I stakeholders to develop a common vision and Pan-European strategy on Embedded Intelligence, that reflects the needs of the industry. In this context, ARTEMIS-ETP is recognized by the European Commission. Smart Mobility with automated vehicles are considered as one of the core activities.

ERTRAC is the European road transport research advisory council. It is the European technology platform (ETP) which brings together road transport stakeholders to develop a common vision for road transport research in Europe. ERTRAC aims at creating and implementing the needed research and innovation strategies for a sustainable and competitive European road transport system. In this context, ERTRAC is recognized and supported by the European Commission.

SafeTRANS is a German not-for-profit competence cluster, comprising member organizations from industry and academia active in the development of electronic components and systems in the transportation domain (cars, planes, trains, ships and their infrastructure). SafeTRANS serves as a communication and knowledge exchange platform for pre-competitive research and development activities. SafeTRANS furthers activities like round tables and theme-oriented working groups, roadmap development, and project incubation, as well as supporting sustainability measures for project results, and providing the link to similar clusters in Europe, to European funding programs and to national public authorities. SafeTRANS’ members are OEMs, suppliers, tool vendors, and system operators as well as research institutes and universities.

A3PS is Austria’s national strategic public–private partnership (PPP) with the goal to support the research and development of advanced propulsion systems and their energy carriers. A3PS enables a close cooperation between its two stakeholders: the bmvit (Austrian ministry for transport, innovation, and technology) and the industry together with research institutions, represented by the members of A3PS.