Thrun et al., Chapter 1

Introduction to Probabilistic Robotics

Why robots must embrace uncertainty instead of fighting it.

Prerequisites: Basic probability (coin flips, conditional probability). That's it.
8
Chapters
1
Simulation
8
Quizzes

Chapter 0: Why Probability?

Imagine you are a robot navigating a hospital corridor. Your laser scanner says the nearest wall is 3.2 meters away. But is it really? The sensor has noise — maybe the true distance is 3.15 or 3.28. Your wheel encoders say you've driven 10 meters since the last turn, but wheels slip on polished floors. Maybe you've only gone 9.7 meters.

Now imagine you are trying to pick up a cup from a table. Your camera detects the cup at pixel coordinates (312, 187). But the camera calibration is imperfect, the lighting creates shadows, and the cup might have moved since your last frame. Every single piece of information the robot receives is uncertain.

Traditional robotics tried to fight this uncertainty. Engineers built better sensors, tighter tolerances, more rigid structures. But uncertainty never goes away completely. Even a $50,000 LIDAR has noise. Even a precision-machined gearbox has backlash. And the world itself is unpredictable — people walk in front of robots, doors open and close, objects get moved.

The core thesis: Instead of fighting uncertainty, embrace it. Represent what the robot knows (and doesn't know) as probability distributions. Use Bayes' rule to update beliefs as new evidence arrives. Make decisions that account for the full range of possibilities, not just a single "best guess."

This is the fundamental insight of probabilistic robotics. A robot that knows it is uncertain is more capable than one that pretends it is not. A robot that maintains a probability distribution over its possible locations can reason about ambiguity. A robot that commits to a single position estimate will eventually make a catastrophic mistake.

Check: Why does traditional (non-probabilistic) robotics eventually fail?

Chapter 1: Five Sources of Uncertainty

Where does all this uncertainty come from? Thrun, Burgard, and Fox identify five fundamental sources that plague every robotic system:

1. Sensor noise
Every measurement is corrupted: laser scatter, camera blur, GPS multipath
2. Actuator noise
Commands don't execute perfectly: wheel slip, motor backlash, wind gusts
3. Model imperfection
Physics models are approximate: friction coefficients, aerodynamics
4. Environment change
The world is dynamic: people walk, doors open, furniture moves
5. Computational limits
Approximations are necessary: discrete grids, finite particles, truncated updates

Each source alone would be manageable. But they compound. A small sensor error leads to a slight mislocalization, which causes the robot to aim its actuators slightly wrong, which accumulates over time. Without a principled way to track this growing uncertainty, the robot's internal model of the world drifts further and further from reality.

Key insight: Uncertainty is not a bug to be eliminated — it is a feature to be modeled. By explicitly tracking how confident the robot is in each piece of information, we can weigh evidence appropriately and make robust decisions even when individual measurements are unreliable.
SourceExampleTypical Magnitude
Sensor noiseLaser rangefinder±1-5 cm per reading
Actuator noiseDifferential drive±2-10% of commanded distance
Model errorFloor frictionCan vary 20-50% across surfaces
EnvironmentPeople in hallwayCompletely unpredictable
ComputationGrid discretizationResolution-dependent
Check: Which of these is NOT one of the five fundamental sources of uncertainty in robotics?

Chapter 2: Traditional vs Probabilistic

To appreciate the probabilistic approach, let's contrast it with what came before. Traditional robotics uses what we might call the best-guess paradigm: at each step, compute the single most likely state, treat it as ground truth, and plan accordingly.

Traditional approach:

• Single "best estimate" of position
• Threshold-based decisions
• Fails silently when estimate is wrong
• Cannot recover from errors
• Brittle in dynamic environments

Probabilistic approach:

• Full distribution over possible positions
• Decisions weigh all possibilities
• Knows when it is uncertain
• Can represent multiple hypotheses
• Gracefully handles ambiguity

Consider a robot in a symmetric hallway. It sees the same pattern of doors on both sides. A best-guess system picks one location and commits. If it picks wrong, every subsequent action is based on a false premise. A probabilistic robot maintains two peaks in its belief distribution — "I might be at position A or position B." As it moves and gathers more evidence, one peak grows and the other shrinks. The robot naturally resolves the ambiguity.

The kidnapped robot problem: Imagine a robot that has confidently localized itself, and you suddenly pick it up and place it somewhere else. A traditional system has no way to notice this happened — it keeps navigating based on its old, now-wrong position. A probabilistic system can detect the discrepancy between expected and actual sensor readings, realize its belief is wrong, and re-localize from scratch. This is the gold standard test for robustness.

The price of the probabilistic approach is computational. Instead of tracking one number (the estimated position), you track a full distribution. But decades of algorithmic advances — Kalman filters, particle filters, sparse representations — have made this practical for real-time robotics. The rest of this book shows you how.

Check: What advantage does maintaining a full probability distribution have over a single best estimate?

Chapter 3: Sensor Noise

Let's make uncertainty concrete. Below is a simulation of a robot taking distance measurements. The robot is exactly 5.0 meters from a wall, but each laser reading is corrupted by Gaussian noise. Click Measure to take readings and watch the histogram of measurements build up.

Noisy Sensor Measurements

True distance = 5.0 m. Sensor noise σ = 0.3 m. Each click takes 10 readings. Watch the bell curve emerge.

0 readings
Noise σ0.30m

Even though each individual reading is imprecise, the collection of readings tells a clear story. The mean converges toward the true value, and the spread reveals the sensor's precision. This is the foundation of probabilistic sensing: no single reading is trustworthy, but many readings, properly combined, are powerful.

Gaussian noise model: Most sensors are well-modeled by a Gaussian (normal) distribution centered on the true value. The standard deviation σ characterizes the sensor's precision. A $100 laser rangefinder might have σ = 5 cm; a $10,000 survey-grade unit might have σ = 2 mm. The math is the same either way.
Check: If a sensor has Gaussian noise with σ = 0.3 m, what does that tell you?

Chapter 4: Actuator Noise

Sensors are not the only source of uncertainty. When a robot commands its wheels to turn 100 rotations, the actual number might be 98 or 103. When a drone commands 2 m/s forward thrust, wind and turbulence mean the actual velocity varies moment to moment. This is actuator noise or motion noise.

The consequences are dramatic over time. Imagine a robot driving in a straight line for 100 meters. Each meter it drives, there's a small random error in both distance and heading. By the end of the 100 meters, the robot might be several meters off course — even though each individual step was "almost correct."

xt = xt-1 + ut + εt,    εt ~ N(0, σ2motion)

Where xt is the actual position, ut is the commanded motion, and εt is the random noise. The variance of the position error grows with every step. After T steps, the total variance is approximately T σ2motion. This is why dead reckoning (navigating by motor commands alone, with no sensor feedback) is doomed to accumulate unbounded error.

Key insight: Motion makes the robot less certain about where it is. Every command adds noise. Sensing makes the robot more certain — every measurement reduces uncertainty. The interplay between these two forces is the heartbeat of the entire book. The Bayes filter (Chapter 2) formalizes this: predict (get less certain), then update (get more certain), repeat forever.
Robot moves
Uncertainty grows (prediction step)
Robot senses
Uncertainty shrinks (update step)
↻ repeat forever

Without periodic sensing to "reset" the uncertainty, a robot quickly loses track of where it is. This is why GPS-denied environments (indoors, underwater, underground) are so challenging — the robot must rely on local sensors (LIDAR, cameras) and clever algorithms to keep its uncertainty from exploding.

Check: What happens to a robot's position uncertainty if it moves but never senses?

Chapter 5: The Big Idea

Here is the central idea of probabilistic robotics, the one sentence you should remember above all else:

The probabilistic approach: Represent the robot's knowledge as a probability distribution over the space of possible states. Update this distribution using Bayes' rule whenever new information (sensor data or motor commands) arrives. Make decisions by reasoning over the full distribution, not just the most likely state.

Everything in this book is an instantiation of this idea. The differences between the algorithms — Kalman filters, particle filters, grid methods — are just different ways to represent and compute with probability distributions. But the underlying logic is always the same: belief → action → prediction → measurement → update → belief.

The mathematical framework is called recursive Bayesian estimation. "Recursive" because we update the belief incrementally with each new piece of data, never needing to reprocess old data. "Bayesian" because the update rule is Bayes' theorem. "Estimation" because we're estimating the hidden state of the world from noisy observations.

bel(xt) = η · p(zt | xt) · ∫ p(xt | ut, xt-1) · bel(xt-1) dxt-1

Don't worry about the details yet — Chapter 2 will derive this step by step. For now, just see the structure: the old belief bel(xt-1) is pushed forward through the motion model, then corrected by the sensor model. The result is the new belief bel(xt). This is the Bayes filter, and every algorithm in this book is a variation on it.

SymbolMeaning
bel(xt)Belief: probability distribution over state at time t
p(zt | xt)Measurement model: how likely this sensor reading given state
p(xt | ut, xt-1)Motion model: how state changes given command
ηNormalizing constant (makes probabilities sum to 1)
Check: What does the Bayes filter use to update the robot's belief?

Chapter 6: Road Map

This book unfolds in a logical sequence. Each chapter builds on the previous ones. Here is how the ideas fit together:

ChaptersTopicWhat you'll learn
1-4FoundationsBayes filter, Kalman filter, particle filter
5-6Sensing & ActingMotion models, sensor models
7-8Localization"Where am I?" — using filters to track position
9Mapping"What does the world look like?" — building maps
10-14SLAM"Where am I AND what does the world look like?" simultaneously
15-16Planning"What should I do?" — decision-making under uncertainty

The three great problems of mobile robotics are localization (where am I?), mapping (what does the world look like?), and planning (what should I do?). In reality, these problems are intertwined — you need a map to localize, you need to be localized to build a map, and you need both to plan. SLAM (Simultaneous Localization and Mapping) tackles the first two together. POMDPs tackle planning under sensing uncertainty.

The filter zoo: Chapters 2-4 present the three major filter families. The Kalman filter (Chapter 3) represents beliefs as Gaussians — fast and elegant, but assumes linearity. The particle filter (Chapter 4) represents beliefs as weighted samples — handles arbitrary distributions, but needs many particles. The histogram filter (Chapter 4) discretizes state space into a grid — simple and flexible, but scales poorly. Understanding the trade-offs between these three is central to the field.
Check: What does SLAM stand for, and why is it hard?

Chapter 7: Summary

This chapter established the philosophical foundation of probabilistic robotics. The key takeaways:

ConceptSummary
Uncertainty is everywhereFive sources: sensors, actuators, models, environment, computation
Probability is the languageRepresent knowledge as distributions, not point estimates
Bayes filter is the enginePredict (motion), update (sensing), repeat
Three filter familiesKalman (Gaussians), particle (samples), histogram (grids)
Three great problemsLocalization, mapping, and planning
What comes next: Chapter 2 formalizes everything. We'll define beliefs mathematically, derive the Bayes filter from first principles, and see it work on a concrete example. The philosophy becomes an algorithm.

Probabilistic robotics has transformed the field. Before these ideas, robots worked in structured factories with known geometry. After, robots navigated hospitals, explored Mars, and drove on highways. The common thread is the willingness to say "I don't know exactly where I am" and work with that uncertainty instead of ignoring it.

"The key idea of probabilistic robotics is to represent
uncertainty explicitly, using the calculus of probability theory."
— Thrun, Burgard & Fox, Chapter 1
Check: What is the fundamental difference between probabilistic robotics and traditional robotics?