Advanced Driver Assistance Systems and Autonomous Vehicles

Advanced driver-assistance systems (ADAS) and autonomous vehicles (AV) have the potential to reshape transportation, by reducing risky driver behaviors, traffic jams, carbon emission, and cost of transportation, as well as improving road safety, independence of seniors and people with disabilities, and human productivity. Although advanced driver-assistance systems and autonomous driving functions are promising, there are many challenges, including new technologies, requalification of non-auto-grade components, and new mission profiles for existing auto-grade components. The chapter reviews various challenges in ADAS and AV, as well as contents of other chapters in the book.

Life-saving advanced driver-assistance systems (ADASs) and autonomous vehicles (AVs) are the fastest growing technology segment in the automotive market. Artificial intelligence (AI) is one of the most critical components in ADAS and AV. Machine learning (ML), deep Learning (DL), simulators, cloud computing, and embedded hardware platforms are entering the equation of ADAS and AV innovation, especially at level four and level five automation, where the classic rule-based ADAS functions reach their limits. This chapter reviews the basic concepts and recent applications of AI in ADAS and AV, including supervised learning, unsupervised learning, reinforcement learning, DL architectures in AVs, mostly used DL algorithms, edge cases and safety, training datasets, simulators, and infrastructures.

The future of driving is quickly evolving toward AI-enabled, fully autonomous vehicles. The centralized Compute system will serve as a nerve center for all autonomous vehicles to meet stringent intelligence, performance, safety, security, and reliability requirements. We’re seeing the complexity of autonomous driving systems growing at an unprecedented rate, and computational processing needs to keep pace with this growth. A high-performance, automotive-grade Compute system must be able to accommodate numerous sensor inputs from cameras, radars, light detection and ranging radars (LiDAR), ultrasonic sensors, inertial sensor module (ISM), acoustic sensors, and Vehicle-to-Vehicle (V2V)/Vehicle-to-Everything (V2X) communications concurrently to accurately and reliably perceive the environment around the vehicles. Also, it must be able to promptly enable better and safer driving decisions including prediction, planning, and control after analyzing all the perceived information. In this chapter, motivations, as well as various, Compute architectures and key components consisting of an advanced autonomous vehicle Compute system such as System on Chip (SoC), memory, storage, and network are reviewed. Furthermore, real-time operating system, onboard management, fault detection and diagnostics, security, and middleware will be illustrated. How to conduct rigid electrical tests and reliability validation to qualify autonomous vehicle Compute will be covered. Finally, challenges in Compute design, manufacturing, and validation including performance, power consumption, thermal management, size, cost, safety, security, quality, and reliability are explored for safe deployment of the autonomous vehicle at scale.

Automotive electronics is among the fastest growing segments of semi-conductor industry (https://​www.​pwc.​com/​gx/​en/​technology/​publications/​assets/​pwc-semiconductor-survey-interactive.​pdf). It is primarily driven by increasing dependence on advanced electronics for a wide range of functions such as safety, control, and driver assist. The critical-to-function Advanced Driver-Assist System (ADAS) must meet stringent reliability and thermal requirements while providing high-performance deterministic compute. This chapter will cover the packaging technologies employed to meet the lifetime performance, quality, and reliability demands of automotive applications, as well as the system packaging and thermal management strategies for control units that dissipate between 10 and 500 W of power depending on the application. The high operating ambient temperatures for passenger vehicles can range from 65 to 85 °C, and this influences the choice of thermal solution ranging from low-power fan-less designs, mid-power forced-air designs to high-power liquid-cooled solutions. While significant innovations in technology for in-vehicle infotainment to autonomous driving are happening today, more are required in the near future to translate into reality the visions of automobile design that will revolutionize the auto industry and more broadly the way we live.

This chapter is focused on the fundamentals of Flash memory and NAND in particular. The explosion of data with the rise of the Internet and mobile computing in the past 30 years has made NAND Flash the most successful nonvolatile memory technology in the world. Compared to other alternatives, the low-cost-per-bit, high Read/Write speed, lightweight and compact form factor, and higher reliability over conventional hard disk drive (HDD) make NAND Flash the storage medium of choice for numerous applications. In the first section of this chapter, the basic floating gate memory cell structure is introduced to illustrate the fundamental physical characteristics that make NAND amenable for device scaling and well suited for data storage applications. Then, the historical evolution of NAND Flash is reviewed. In the second section, NAND fundamentals, including basic operations, memory architecture, and manufacturing processes, etc., will be discussed in detail. Starting around 2014, the NAND industry went through a major architecture transition from 2D to 3D NAND. The third section is devoted to the unique technology and design challenges of 3D NAND. It is followed by a discussion of various reliability issues, before the final conclusion on the 3D NAND future outlook.

Interconnect could provide physical and logical connection between two electronic devices with interconnect with affordable quality and reliability, and thus interconnect is critical for the development of Advanced Driver Assistance Systems (ADAS). In this chapter, interconnects and solder joint technology for applications under the hood are reviewed. In addition, the bonding techniques by using Cu-Cu direct bonding and hybrid bonding are presented.

Cameras have become one of the most important sensors in advanced driver-assistance systems (ADAS) and autonomous driving (AD) vehicles. There are different ways to categorize cameras in ADAS/AD vehicles based on the camera’s placement, application, and technology. Most camera systems consist of hardware components that compose the camera module and image processing components that control the hardware and perform digital operations on captured images. An overview of camera systems, hardware, image processing, and product development processes is introduced in this chapter.

In the 2005 DARPA Grand Challenge, Stanford’s Stanley robot car, navigated by 5 roof-mounted SICK Lidars, won the race against 4 other competitors. Since then, Lidars have gradually become crucial perception sensors for autonomous driving and ADAS, due to their ability to generate real-time point clouds with accurate 3D information of the vehicle’s surroundings. Extensive research efforts have been invested into Lidar technology in both academia and the industry. As a result, a diverse variety of Lidar sensors have been created in the past decade. In this chapter, the authors aim to review the state of the art of Lidar sensors for autonomous driving or ADAS applications. The manuscript discusses the important metrics for Lidar sensor performance: detection range, field of view (FOV), angular resolution, frame rate, and eye safety. Then, different Lidar mapping methods and distance calculation mechanisms are discussed. Current status of mechanical, MEMS, FLASH, optical phased array (OPA), and frequency-modulated continuous wave (FMCW) Lidars is introduced, and their pros and cons and reliability performance are compared.

Over the last decade, new applications have emerged for radar technology in the automotive industry, such as adaptive cruise control, blind spot detection and automatic emergency braking. It is expected that the continued development of radar will unlock new capabilities for autonomous vehicles and safety systems. Indeed, advancements in integrated circuit design have enabled the development of very high-frequency radars with sophisticated signal processing and machine learning techniques. As a consequence, it is now possible with low power, small form factor and low-cost sensors to have rich point cloud data in dense environments. However, many challenges remain that can reduce precision, recall and accuracy of automotive radar, intra- and inter-platform interference or jamming, multi-radar fusion, multipath effects, false targets and detection ambiguity (range, velocity and angle). Many approaches have been developed to address these challenges; massive hybrid MIMO (Multiple-Input Multiple-Output), compressed sensing methods, sparse arrays, PMCW (Phase-Modulated Continuous Wave) Radar, Artificial Intelligence (AI) algorithms and others. Some of these approaches have already been applied to automotive radars, and some likely will be used in the coming years. This chapter will review why radar is unique, the challenges and the state-of-the-art solutions. We will use intuitive and simplified examples to provide the background needed to understand the utility of radar and these techniques.

Battery packs have been widely used as the main power source for advanced driver-assistant vehicles. The status and challenges related to electrochemical batteries, including material choices, energy density, performance, battery design, safety, reliability, cost, and development trend, are reviewed and addressed in this chapter. Meanwhile, other types of the electric power sources, such as fuel cells and capacitors, and challenges related to power management systems, are briefly introduced and discussed.

Visualization technologies are the most vital components of in-vehicle interactions. The shift toward autonomous vehicles and connected cars is bringing a future in which occupants would be needed to monitor the status of the vehicle and its surroundings. Meanwhile, occupants would also spend significantly more time watching displays for entertainment, information, and connectivity on the road. Therefore, the need for in-vehicle displays with better visibility, brightness, viewing angle, resolution, sharpness, and reliability together with larger size and free-form that offer unobtrusive visual information during journeys is on the rise. Superior display with touch technologies can enable a safe, informative, and comfortable driving or riding experience. The applied in-vehicle display products include center infotainment display, rear-seat entertainment display, head-up display, side mirror display, and instrument cluster display. In this chapter, motivations, as well as various architectures of display including LED, LCD, OLED, mini-/micro-LED, TFT, flexible, head-up display, and touch screen, will be introduced. Designing displays into vehicles imposes very different challenges than designing them for consumer applications. This is due to some unique factors associated with vehicle usages, such as the required product life cycles, the extremely harsh environment, frequent mechanical impacts, the stringent EMI/EMC compliance, the required high-level ESD protection, and functional safety requirements. Requirements and challenges of display in-vehicle application, including fabrication, characterization, inspection, quality, reliability, EMI/EMC/ESD, and failure analysis are reviewed.

In the modern era of the Internet of Things (IoT), data generation speed is exploding. Autonomous vehicles (AV), one type of IoT, are generating enormous amounts of data during operation. Most of the data will be saved in the data center. As the major data storage in data centers, hard disk drives (HDD) have a history of more than 60 years. HDD is the art of combining magnetic sensors, electromechanical components, and electronics. To meet the increasing demand of data storage, next generation HDD is under development with tremendous advantage in data areal density for high-capacity data storage, but meanwhile with challenge in reliability.

The role and responsibility of reliability engineers change with project milestones. During the design phase, reliability engineers define reliability targets, lead teams to review design weakness, brainstorm potential failure mode and root cause, define test plan, customize stress profile, allocate samples, and prepare for test program and test equipment. During the development phase, reliability engineers execute the test, analyze data by fitting life distribution and doing hypothesis tests, and drive failure analysis and corrective action. Once a product is released to market, reliability engineers work on field return and warranty analysis. This chapter covers some of these topics, including risk assessment methodologies (failure mode and effect analysis, fault tree analysis and stress-strength analysis), accelerated life testing and highly accelerated life testing, reliability statistics (sample size calculation, life distribution analysis by Linear least square regression and Maximum likelihood estimation, confidence interval calculation, hypothesis tests for mean and variance), failure analysis and corrective/preventive actions, system reliability metrics, reliability block diagram methods, and repairable system. Various case studies are used to illustrate the ideas, including cameras, cold plates, dash mount audio device, LED display, Lidar bracket, magnetic sensor, network and multimedia PCB boards, power supplies, Radar, and waterblock.

Failure analysis (FA) could provide timely feedback to process optimization and solution paths for system failures; thus, it is critical for the development of advanced driver assistance systems (ADAS). In this chapter, failure analysis flows starting from systems or boards until components or packages and dice are introduced. Electrical fault isolation (FI) techniques designed to locate subtle defects inside complicated semiconductor devices are reviewed. Physical failure analysis approaches adopted to provide artifact free nanometer scale analysis are discussed. Material analysis methods assisting in thorough root cause investigation are presented. Non-destructive and high-resolution imaging tools with the potential of significantly shortening failure analysis through put time are demonstrated. Case studies are used to illustrate strategies and methodologies in ADAS failure analysis.

Ball bonding is the most widely used interconnection method in microelectronic packages. It has enabled many modern technologies including medical implants, aerospace, automobiles and Internet of Things. In the automotive industry, driving automation and advanced driver assistance systems motivated mainly by safety enhancement are gaining traction. The reliabilities of these technologies necessitate those of the underlying ball bonds. This chapter provides interpretation of the mechanisms of corrosion that causes reliability failures of Cu and Au ball bonds, by unifying approaches based on microstructure characterization and electrochemical investigation. The corrosion of Cu ball bonds starts with pitting of the most Cu-rich layer (MCRL) under the chlorinated water layer, evolves into crevice corrosion, and can be assisted by stress corrosion cracking. In the MCRL, Al is preferentially oxidized, while the Cu atoms remain largely immune and coalesce to form nanoparticles. Four methods to address the corrosion are presented. Limited data indicate the same corrosion mechanisms for Au ball bonds.