- Shopping Bag ( 0 items )
Revised and updated, the second edition includes several new chapters with projects and applications. The authors keep pace with the ever-growing and rapidly expanding field of robotics. The new edition reflects technological developments and includes programs and activities for robot enthusiasts. Using photographs, illustrations, and informative text, Mobile Robots guides the reader through the step-by-step process of constructing two different and inexpensive yet fully functional robots.
Robotics is about building systems. Locomotion actuators, manipulators, control systems, sensor suites, efficient power supplies, well-engineered software -- all of these subsystems have to be designed to fit together into an appropriate package suitable for carrying out the robot's task. Where do we start?
We think of a robot as an intelligent connection of perception to action. The implementation of that goal might take on a variety of "costumes," from mechanical logic to microprocessor control to networks of neuron-like gates. Our approach is to create abstraction barriers in terms of thinking about the intelligent capabilities our robot might possess and then to gradually break them down by explaining the specific hardware details that we might employ to create those competences. The theme throughout is to build systems early and build systems often -- to start with very simple systems that connect perception to action and to gradually move to more sophisticated machines.
We start with a tutorial in the next chapter that describes how to build a robot, TuteBot, that is able to wander around a room and avoid obstacles. This example robot, pictured in Figure 1.1, is implemented without recourse to a microprocessor. TuteBot is merely an agglomeration of switches, relays, motors, and discrete electronic components, all of which can be assembled rather easily. You will be able to adjust TuteBot's reflexes by tweaking two potentiometers.
From this very simple example of a robot, we introduce the microprocessor and the advantages of using software to manage the complexity of large numbers of sensors and actuators. The viewpoint from this moment on is to build systems with the intent of getting to software as soon as possible. To keep parts count, size, and costs down for our readers, we describe minimalist ways to interface sensors, motors, and power supplies in another example robot, Rug Warrior. The microprocessor becomes the heart of Rug Warrior, and the following chapters describe the workings of mechanical and electrical components and the interface circuitry that enables them to be driven from a microprocessor. Software-primitive operations are threaded throughout the book as each new perception or locomotion system is introduced.
Although this book describes the details involved in actually building robots, we hope also to raise some deeper points about models of intelligence. What is intelligence? Is it the contemplative thought involved in playing chess? Is it the reflexive action that occurs as you try to keep the gnats out of your eyes while walking down the street on a hot, muggy summer night? Or is it the common-sense reasoning used in deciding what to make for breakfast? We will stick with the notion that intelligence is the foundation for how people act most of the time. It will be interesting to keep some of these questions in mind as we investigate the sorts of mechanisms we can use to endow our example robots with low-level behaviors.
Other features of intelligence have to do with the role the environment plays in our view of cleverness. How connected are sensing and actuation to intelligence? How much of what we acknowledge as complex behavior is merely a reflection of simple behaviors off of a complex environment? For instance, if we observe the behavior of ants scurrying around their anthills, we might begin to wonder whether their complex paths result from careful planning and deep contemplation, or perhaps merely from simple rules of behavior acted out in an environment full of uneven terrain, obstacles to climb over and other ants.
TuteBot and Rug Warrior will not answer many of these questions pertaining to the structure of intelligence, but we hope that they can be the platforms for an inexpensive, easily attainable AI input/output device -- a collection of sensors and actuators that provide a little bit of input, a little bit of output, and a little bit of computation to readers interested in experimenting with some of these issues.
Many of the modern theories in artificial intelligence grew from work in a number of other fields. Cybernetics, in the 1940s and 1950s, was a field of research that tried to understand intelligence through the study of the control of machines. Cybernetics developed in parallel with classical control theory. Its model of computation was analog, and it tried also to understand intelligence in animals by modeling them as machines. Our example of TuteBot is very much in the same spirit as the early work in cybernetics.
For instance, Figure 1.2 illustrates the extent of TuteBot's talents. The long dashed lines at the bottom of the figure exemplify one initial behavior, where TuteBot moves forward in a straight line until it hits an obstacle. It then backs up, turning left for some period, and then proceeds forward again in a straight-line motion.
A number of mechanisms could be imagined necessary to achieve this behavior. We could suggest contemplative recognition of chair legs and walls and TuteBot making explicit decisions concerning when to back up and how far to turn, but TuteBot has no such model of the world. Instead, TuteBot has a simple analog electrical circuit for a control system, which directs TuteBot's two wheels to move it forward until a bump sensor on the front detects a collision. The signal from the bump sensor directs both motors to reverse direction, and TuteBot then backs up. What makes it turn is an element of state, or timing, in the system that is implemented with a resistor-capacitor (RC) circuit, one for each wheel. If the RC circuit on each wheel is set differently, one wheel will back up for a longer period of time than the other wheel, causing TuteBot to turn. When TuteBot resumes forward motion, it no longer has the same heading and so avoids ramming the obstacle it first bumped.
A second behavior can be added to TuteBot using a similar strategy. ...
|Preface to the Second Edition|
|4||Designing and Prototyping||95|
|12||Robot Design Principles||355|
|B||Rug Warrior Programs||377|
|G||Color and ASCII Codes||441|
Posted December 15, 2002
This is a good book on robotics, but I would not recommend this book to the beginner who has no idea in this field. An accompanying beginner's book in robotics or electronics should be a good combination with this book.Was this review helpful? Yes NoThank you for your feedback. Report this reviewThank you, this review has been flagged.