# Foundations of Autonomous Driving: How Self-Driving Vehicles Perceive, Think, and Move Canonical URL: https://cretisoftbooks.com/en/books/foundations-of-autonomous-driving-for-engineers Book page: https://cretisoftbooks.com/en/books/foundations-of-autonomous-driving-for-engineers Author: Neil Carden Language: en Description: Imagine a late afternoon on a city street: sunlight glares off a windshield, a delivery van partially hides a pedestrian, and the driver ahead brakes for no reason a machine can easily recognize. This is the kind of ambiguous, high-stakes moment that every self-driving vehicle must handle safely, and it happens thousands of times during an ordinary test drive. It is not enough to have a large dataset or a high-resolution camera; the car must sense the world through multiple physical modalities, fuse them into a coherent model, reason about what it sees in a probabilistic way, predict the immediate future of every moving actor, and convert that judgment into a smooth steering and throttle command. The entire autonomous driving stack must operate in concert, and the goal of this book is to make that stack transparent—layer by layer, from the physical sensors to the safety cases used for deployment. Foundations of Autonomous Driving: How Self-Driving Vehicles Perceive, Think, and Move is a comprehensive, vendor-neutral guide to the technology behind driverless cars. Unlike a collection of isolated algorithms, this book structures the entire discipline around the Sense–Think–Act cycle, which acts as a single mental model for understanding every component. It takes you from the SAE levels of automation and the operational design domain all the way to the software stack that runs on the vehicle. The book is written for practitioners and enthusiasts who want to move beyond marketing materials and truly understand how autonomous vehicles work. It is a technical book, but it assumes no deep background in machine learning or advanced calculus; each idea is introduced through physical intuition, system block diagrams, and simplified mathematics. With a target of about five hundred pages, it balances depth and accessibility, making it useful both as a structured read and as a reference. The book begins at the physical layer, because every decision an autonomous vehicle makes flows from the hardware in the car. You will explore the transition from mechanical systems to drive-by-wire, which replaces steering columns and brake lines with electronic signals and actuators. A dedicated chapter maps the main electronic control units, central computers, and automotive networks such as CAN and Automotive Ethernet, showing how sensor data travels through the vehicle. From there, the text spends two full chapters on the primary sensors: camera systems and their limitations in dynamic range, radar and its FMCW principle that works through rain and spray, and LiDAR with its scanning mechanisms and dense point clouds. The book compares their strengths and trade-offs in real driving conditions and explains how placement affects coverage. It also covers the supporting sensors—GNSS, IMU, wheel odometry, ultrasonic—and the often-underappreciated problems of calibration and time synchronization. Sensor fusion then becomes the unifying thread, with a clear explanation of coordinate frames, data association, and the intuition behind probabilistic fusion methods like the Kalman filter, so you understand how a multi-sensor suite becomes a dependable world representation. The core of the book is the intelligence layer, where raw sensor data becomes a structured understanding of the world. You'll learn how deep neural networks are used for object detection, multi-object tracking, lane detection, traffic sign and light recognition, and drivable-space estimation. Beyond 2D detection, the text explains depth estimation, 3D bounding boxes, bird's-eye-view representations, and occupancy grids, which are necessary for safe motion planning. A major theme is localization: how the vehicle determines where it is by combining GNSS with RTK corrections, visual odometry, LiDAR scan matching, and high-definition maps, and how simultaneous localization and mapping (SLAM) makes operation possible without prior maps. The prediction chapters focus on anticipating what other road users will do—physics-based motion models, maneuver recognition, and uncertainty-aware multi-agent interaction. This section is carefully built to show the trade-offs between accuracy, computational cost, and robustness, because those trade-offs define the real engineering of autonomous driving. The final technical part of the book turns to decision making and execution, then zooms out to the systems engineering required for real-world deployment. You'll see how route planning, behavior planning, path planning, and motion planning form a hierarchy: the vehicle decides which road to take, which maneuver to execute, and which precise trajectory to follow, all while respecting driving rules and avoiding collisions. The control chapter clarifies how planned trajectories are tracked with longitudinal and lateral control, explaining PID, pure pursuit, the Stanley controller, and the intuition behind model predictive control. Safety is treated as a design discipline, not an afterthought: the book covers redundant sensor and compute architectures, fault detection, fail-operational strategies, difficult weather and edge cases, and the essentials of ISO 26262 functional safety. You'll also learn how autonomous vehicles are architected in software, with modular stacks, publish-subscribe middleware, real-time processing constraints, and data pipelines, including practical insights from open-source frameworks like ROS 2, Autoware, and Apollo. Finally, validation is presented as a layered pyramid of simulation, software-in-the-loop, hardware-in-the-loop, scenario-based testing, closed-course testing, and supervised public-road testing, showing how the industry builds confidence before a robot car meets the public. This book is for anyone who needs a reliable mental model of autonomous driving. Software engineers moving into robotics will find the architecture clearly mapped; technical product managers will gain the language to evaluate self-driving projects; engineering students will see how methods from computer vision, control theory, and systems engineering fit together; and automotive professionals will understand the constraints that shape the hardware. The prerequisites are modest: basic programming logic, high-school physics, and a willingness to reason about uncertainty. No advanced calculus or top-level machine learning background is needed, because the text introduces every concept with physical intuition and a simplified mathematical form. By the end, you will be able to trace the path from a photon hitting the camera lens to an actuator turning the front wheels, and you will understand why each module in the pipeline behaves the way it does. More importantly, you will have the vocabulary and conceptual tools to approach the research literature, open-source codebases, and even the safety cases of a modern autonomous vehicle with genuine confidence. AI summary: Foundations of Autonomous Driving offers a comprehensive, vendor-neutral technical introduction to self-driving vehicle technology. The book follows the Sense-Think-Act cycle to explain the complete AV stack, from camera/radar/LiDAR sensors and sensor fusion to deep learning-based perception, localization and mapping, prediction, planning, control, and system safety. It is designed for software engineers, engineering students, technical product managers, and automotive professionals who want a rigorous but accessible understanding of how autonomous vehicles perceive, think, and move. Target audience: Software engineers, engineering students, technical product managers, and automotive professionals seeking a rigorous yet accessible understanding of autonomous vehicle technology. Audience persona: A software engineer or engineering student who wants a comprehensive, systems-level understanding of self-driving car technology, from sensors to control, without deep mathematical prerequisites. Search intent: People searching for a technically grounded, vendor-neutral introduction to how autonomous vehicles perceive, plan, and move, covering the full software and hardware stack. Unique angle: This book uses a unified Sense-Think-Act framework to explain the entire autonomous driving stack in a vendor-neutral, systems-engineering approach, making advanced AV concepts accessible to engineers without deep backgrounds in machine learning or calculus. Content type: technical engineering guide Answer snippets: - An autonomous vehicle uses a Sense-Think-Act cycle: sensors perceive the environment, algorithms reason about the situation, and controllers execute steering and speed commands. - Self-driving cars combine cameras, radar, and LiDAR with GNSS, IMU, and wheel odometry to create a robust, redundant view of the world. - The book explains the full AV software stack, including deep learning object detection, 3D scene understanding, localization with HD maps, trajectory planning, and model predictive control. - Safety is handled through redundancies, fail-operational design, ISO 26262 functional safety, and extensive testing from simulation to public-road pilots. Key topics: Sense-Think-Act architecture, SAE levels of automation, Operational Design Domain, Automotive sensors (camera, radar, LiDAR), Sensor fusion, Perception and object detection, Localization and HD maps, Prediction and planning, Vehicle control, Safety and testing Entities: Sense-Think-Act, SAE J3016, Operational Design Domain, LiDAR, FMCW radar, Sensor fusion, Kalman filter, HD maps, SLAM, PID control, Model Predictive Control, ISO 26262 Problems solved: - Understand how a self-driving car perceives its environment using multiple sensors. - Learn how sensor fusion combines camera, radar, and LiDAR to build a robust world model. - Grasp the planning and decision-making hierarchy from route planning to motion planning. - Discover how control algorithms track planned trajectories. - Understand the safety and validation processes required for real-world deployment. Who should read: - Software engineers transitioning into autonomous driving - Engineering students in robotics, AI, or automotive fields - Technical product managers overseeing AV projects - Automotive professionals seeking to understand the full stack - Robotics enthusiasts who want a structured introduction to AV technology Who should not read: - Readers looking for a non-technical overview of self-driving cars - Those expecting hands-on code tutorials or implementation details - People seeking deep mathematical derivations of every algorithm - Readers wanting a focus on a specific company's autonomous driving solution FAQ: Q: What is the Sense-Think-Act cycle in autonomous driving? A: It is a high-level architecture that organizes an AV's software stack into sensing the environment, thinking about a safe trajectory, and acting through steering and acceleration commands. Q: What sensors do self-driving cars use? A: Autonomous vehicles typically use cameras, radar, and LiDAR for exteroception, along with GNSS, IMU, wheel odometry, and ultrasonic sensors for positioning and proximity detection. Q: What are the SAE levels of automation? A: SAE J3016 defines six levels from Level 0 (no automation) to Level 5 (full automation under all conditions). Levels 1-2 are driver assistance; Levels 3-5 involve increasing vehicle responsibility. Q: How does sensor fusion improve autonomous driving? A: Sensor fusion combines complementary strengths of cameras, radar, and LiDAR, using techniques like Kalman filtering to produce a robust, redundant world model that outperforms any single sensor. Q: What safety standards apply to autonomous vehicles? A: The book introduces ISO 26262 functional safety, which defines Automotive Safety Integrity Levels (ASILs) for electronic systems, along with testing practices like SIL, HIL, scenario-based testing, and closed-course trials. SEO keywords: autonomous driving for engineers, how self-driving cars work, self-driving car technology explained, autonomous vehicle perception and planning, sensor fusion for autonomous vehicles, autonomous driving software stack, robotics for software engineers, self-driving car safety and testing, AI for autonomous vehicles, vehicle control and planning Table of contents: - How to Use This Book - Foundations - Introduction to Autonomous Driving - What Is Autonomous Driving? - ADAS vs Autonomous Driving - SAE Levels of Automation - Operational Design Domain - Applications and Challenges - How an Autonomous Vehicle Works - Sense–Think–Act - The Autonomous Driving Stack - Data Flow Through the System - Hardware and Software Integration - Vehicle and Sensor Architecture - Autonomous Vehicle Hardware - Vehicle Architecture - ECUs and Central Computing - Drive-by-Wire - Steering, Braking, and Throttle - Automotive Networks - Cameras, Radar, and LiDAR - Camera Systems - Automotive Radar - LiDAR - Strengths and Limitations - Sensor Placement - Positioning and Supporting Sensors - GNSS - IMU - Wheel Odometry - Ultrasonic Sensors - Sensor Calibration - Time Synchronization - Sensor Fusion Fundamentals - Why Sensor Fusion Matters - Coordinate Systems - Combining Sensor Measurements - Camera–Radar Fusion - Camera–LiDAR Fusion - Introduction to Probabilistic Fusion - Understanding the Environment - Perception - Object Detection - Object Tracking - Lane Detection - Traffic Signs and Lights - Drivable Space - 3D Scene Understanding - Depth Estimation - 3D Object Detection - Bird’s-Eye View - Occupancy Representation - Environmental Models - Localization and Mapping - GNSS Localization - Visual Localization - LiDAR Localization - HD Maps - Introduction to SLAM - Map-Based and Mapless Driving - Driving Intelligence - Prediction - Motion Prediction - Intention Prediction - Vehicles, Pedestrians, and Cyclists - Multi-Agent Interaction - Prediction Uncertainty - Planning and Decision Making - Route Planning - Behavior Planning - Path Planning - Motion Planning - Collision Avoidance - Driving Rules and Interactions - Vehicle Control - Longitudinal and Lateral Control - Steering and Speed Control - PID Control - Pure Pursuit - Stanley Controller Sample EPUB: https://cretisoftbooks.com/book-samples/6a75ea9b95d831b42e2f3b34-1786265276157-foundations-of-autonomous-driving-how-self-driving-vehicles-perceive-think-and-move-epub-mau-20.epub Purchase links: - Google Books: https://play.google.com/store/books/details?id=1ej_EQAAQBAJ