No. 173 of 500 No. 30 in Machine Learning

Probabilistic Robotics

By Sebastian Thrun, Wolfram Burgard and Dieter Fox

Print length
672 pages
Publisher
The MIT Press
Reading level
Advanced
ISBN-13
978-0262201629
Published
August 19, 2005
Written for
Graduate students, robotics engineers and autonomous-vehicle developers working on state estimation, mapping and planning.

Synopsis

Probabilistic Robotics starts from a simple premise: a robot never knows exactly where it is or what surrounds it, so it should represent that uncertainty explicitly and reason with it. Sebastian Thrun, Wolfram Burgard and Dieter Fox organize the whole subject around one mathematical foundation, recursive Bayesian state estimation, and then derive its main tools: Kalman and extended Kalman filters, histogram and particle filters, and probabilistic models of robot motion and sensor readings. Each chapter pairs derivations with pseudocode, a practitioner's discussion and exercises.

The book then applies those tools to the field's core problems. Two chapters cover localization, including Monte Carlo localization, which the authors helped develop. Five chapters address mapping, from occupancy grids to simultaneous localization and mapping through EKF SLAM, GraphSLAM, the sparse extended information filter and FastSLAM. A final section treats planning and control with Markov decision processes, partially observable MDPs and exploration. The material predates deep learning, so modern perception networks are absent, but the state-estimation methods remain widely used in autonomous vehicles, drones and mobile robots, and readers gain the vocabulary needed to read current robotics research.

Scorecard

Editorial score

How it scored

Editorial score: 81.6 / 1005 weighted criteria

  • Influence30%

    27.0/30
  • Authority25%

    23.8/25
  • Practical value20%

    15.0/20
  • Currency15%

    9.8/15
  • Readability10%

    6.0/10
  • Editorial score

    81.6/100

How the rank was computed

Rank No. 173 of 500 comes from a composite of 57.9: observable signals 21.1 of 55 points plus editorial review 36.8 of 45 points.

All 11 inputs
InputPoints
Observable signals21.1/55
Reach (Amazon ratings volume)8.3/16.5
Consensus across best-of lists5.5/13.8
Awards and bestseller evidence0.0/8.3
Reader rating, sample-adjusted6.4/8.3
Amazon sales rank0.9/5.5
Recency0.0/2.8
Editorial review36.8/45
Author authority10.7/11.3
Influence12.2/13.5
Practical value6.8/9.0
Readability2.7/4.5
Currency4.4/6.8

Each criterion is scored from 1 to 10 and weighted, and the points add up to the editorial score. The rank combines observable signals (55 percent) with this score (45 percent). Read the methodology.

Verdict

The standard graduate text on robot perception and control under uncertainty, still foundational for SLAM, localization and autonomous vehicles.

Strengths

  • Unified Bayesian treatment of robotics
  • Pseudocode, derivations and exercises throughout
  • Written by the field's leading researchers

Keep in mind

It assumes solid probability and linear algebra, predates deep learning-based perception, and offers pseudocode rather than runnable code for modern libraries.

Key takeaways

  1. Robots should maintain probability distributions over their state, not single best guesses, so that noisy sensors and imperfect motion are handled systematically.
  2. The Bayes filter unifies the field; Kalman, particle and histogram filters are different ways of approximating the same recursive update.
  3. Simultaneous localization and mapping can be solved with several families of algorithms, each trading accuracy, scalability and computational cost differently.
  4. Planning under uncertainty, formalized as POMDPs, lets a robot choose actions that gather information as well as actions that reach goals.

Notable facts

  • Lead author Sebastian Thrun led the Stanford Racing team whose vehicle Stanley won the 2005 DARPA Grand Challenge, and the book is widely used and translated into multiple languages. Source (opens in a new tab)
  • Electronic Design's review judged it invaluable for robotics readers with a strong probability background and said it could fill two graduate semesters. Source (opens in a new tab)

About the author

Sebastian Thrun, Wolfram Burgard and Dieter Fox

Sebastian Thrun led the Stanford team whose car Stanley won the 2005 DARPA Grand Challenge, co-founded Google X and its self-driving car project, and co-founded Udacity; he is a National Academy of Engineering member. Coauthor Wolfram Burgard, a Leibniz Prize winner, leads robotics and AI at the University of Technology Nuremberg; Dieter Fox is a University of Washington professor who directed robotics research at NVIDIA.

Questions

Is Probabilistic Robotics still relevant in 2026?
Yes for its core methods. Bayes filters, particle filters, Kalman filtering and SLAM remain widely used in autonomous vehicles, drones and mobile robots. The 2005 text does not cover deep learning for perception or learned end-to-end control, so readers should pair it with current papers or newer courses on robot learning.
What background do I need for Probabilistic Robotics?
Plan on comfort with probability, linear algebra and calculus, plus some programming experience to implement the pseudocode. The book is pitched at graduate students and practitioners; Electronic Design's reviewer called it far from light reading, with enough material to fill two graduate semesters. Undergraduates may want an introductory robotics course first.
What topics does Probabilistic Robotics cover?
It is organized in four parts: basics, covering recursive state estimation, Gaussian and nonparametric filters, and motion and sensor models; localization, including Monte Carlo localization; mapping, from occupancy grids to SLAM methods such as GraphSLAM and FastSLAM; and planning and control, covering Markov decision processes, POMDPs and exploration.

Sources

As an Amazon Associate, Best AI Books earns from qualifying purchases. Best AI Books is sponsored by Iternal.ai, whose founder and CEO, John Byron Hanby IV, wrote The AI Strategy Blueprint, our #1 ranked book. How we rank books.