“you just can't differentiate between a robot and the very best of humans.”
― Isaac Asimov, I, Robot
“One man’s “magic” is another man’s engineering.”
― Robert Heinlein
This is not a beginner’s robotics book.
And it is absolutely not for hobbyists.
This book was written for engineers, researchers, and advanced practitioners who already understand one hard truth:
Robotic systems don’t fail because of missing algorithms — they fail because assumptions go untested.
Most robotics books show you what to build.
Very few show you how systems break, why results don’t reproduce, or how to prove performance before hardware ever moves.
That is exactly what this book does.
Why simulation matters (and why most people misuse it)
Real robots are expensive, noisy, and misleading.
Simulation, when used correctly, is the opposite — it exposes weaknesses clearly and repeatably.
Volume 1 treats simulation as a scientific instrument, not a demo environment.
You’ll learn how to:
- Design perception, planning, control, and learning systems that survive stress
- Quantify uncertainty instead of ignoring it
- Detect drift, instability, and divergence early
- Run controlled experiments where every assumption is visible
This is robotics as systems engineering, not optimism.
What this book actually covers
Every topic is treated as an experiment, not a tutorial:
- Visual–inertial SLAM with noise, degeneracy, and drift exposed
- EKF vs UKF under real uncertainty
- Sampling-based and optimization-based planning at scale
- MPC, nonlinear control, and hybrid systems under failure conditions
- Reinforcement learning examined for instability and overfitting
- Multi-robot coordination pushed to communication limits
- Safety, fault injection, adversarial attacks, and verification
- Physics engines, numerical stability, and simulator bias
- Large-scale parallel simulation and reproducible benchmarking
If something fails, the book shows how to measure it and why it failed.
Who this book is for
This book is for:
- Robotics engineers working on real systems
- Graduate students and researchers who need reproducible results
- Professionals building autonomy stacks, not demos
It assumes comfort with math, control theory, probability, and robotics software.
If you want shortcuts, this book will frustrate you.
If you want clarity, it will save you years.
Why Volume 1 matters — even if you “know robotics”
Most failures happen between disciplines:
- Estimation breaking control
- Learning exploiting simulator artifacts
- Multi-robot systems collapsing under scale
Volume 1 forces these interactions into the open.
You’ll finish the book thinking less in modules — and more in coupled dynamical systems.
About the two-volume structure
Volume 1 focuses on rigorous simulation-based design, analysis, and verification.
Volume 2 moves into interaction, embodiment, and large-scale simulated worlds.
Volume 1 gives you the discipline required to do Volume 2 correctly.
Final word
This is a serious engineering text.
It does not motivate.
It does not simplify.
It does not pretend robotics is easy.
It teaches you how to think clearly, test honestly, and fail safely — before hardware does.
If that’s what you want,
Experimental Robotics Projects — Volume 1 belongs on your desk.
