This book has all of the right topics: control, finite state machines, search algorithms, neural networks, etc. However, in the end, you really won't learn anything. It is ponderous that this author could fill so many pages with text, figures, equations, and code and never actually tell you anything. The problem is, that everything he tells you is correct. However, in the end, it is like a jigsaw puzzle with all of the pieces thrown up in the air. Yes, he shows Java code, but the guts of the algorithms that make it perform artificial intelligence tasks are missing. Yes, his equations are all correct, but he doesn't tell you how he is going to take these equations, form an algorithm, and then turn that into Java code that works to control a robot - not on a macro level, not on a micro level.
I suggest that instead of this book you look at several other good ones on the subject. For learning neural networks and their implementation in Java, I really enjoyed "Introduction to Neural Networks with Java" by Heaton. For a good practical book on behavior-based robotics try "Robot Programming : A Practical Guide to Behavior-Based Robotics" by Jones and Roth. They do an excellent job of describing state machines and algorithms for robotic control where this book fails miserably. "AI Application Programming" does a good job of explaining all of the AI algorithms on a practical level with plenty of working code. Finally, you might want to look at "The Definitive Guide to Building Java Robots" only if you already know the algorithms involved in building robots, but you would like a tutorial in how some of the specialized Java API's might help you perform this task. For example it shows how to use the Java Speech API implementation, FreeTTS, for doing speech recognition and synthesis on a robot.