slam-handbook-contributors/slam-handbook-public-release

This repository holds the public release of the SLAM Handbook, a comprehensive academic textbook on Simultaneous Localization and Mapping that is being published by Cambridge University Press. SLAM is the technology that lets a robot, drone, or autonomous vehicle figure out where it is in the world while at the same time building a map of its surroundings, without GPS or any pre-existing map to rely on. It is a foundational problem in robotics and is used in self-driving cars, delivery drones, Mars rovers, and augmented reality headsets. The book was written by a large group of academic experts in the field. It is structured in three parts released incrementally: Part 1 covers the theoretical foundations, Part 2 covers practical SLAM systems organized by sensor type, and Part 3 covers the future of the field and its relationship to spatial AI. The first part was released in November 2024, the second in March 2025, and the third in May 2025. The chapters in Part 1 address topics like factor graphs (a mathematical framework for organizing sensor data), handling bad sensor readings and outliers, and different ways to represent maps in memory. Part 2 covers SLAM systems built around specific sensors: cameras, LiDAR laser scanners, radar, event cameras, and inertial measurement units. Part 3 looks at learned and deep SLAM approaches, multi-robot SLAM, and what it would mean for a robot to have genuine spatial intelligence. The repository does not contain software you run. It is a distribution channel for the book itself. The README consists primarily of BibTeX citation entries, one per chapter, so that researchers can properly cite individual chapters in their own papers. Questions, corrections, and suggestions can be submitted through the GitHub Issues and Discussions tabs on this page.