Robotics Books

The automation of trading and negotiation is becoming more important with each passing year in the twenty-first century. The New York Stock Exchange, the Chicago Board of Trade, NASDAQ, and Web-based businesses like Amazon and Ebay are facing pressure to automate not only their purchasing, but also their abilities to negotiate business deals on part of their customers. In addition, recently issues with electricity and bandwidth marketing will entail a reliable and efficient method of automated bargaining in order to ensure reliable electrical grids and efficient bandwidth allocations. The automation in all of these areas will need to be trustworthy, efficient, and profitable for both the customer and the company if it is to be a viable part of business life.

This book outlines a mathematical theory of negotiation, and combines concepts from game theory, mathematical economics, and artificial intelligence in order to build strategic negotiation models, and discusses their empirical validation. The author bases her model on the Rubenstein model of alternating offers, wherein intelligent agents exchange offers until they reach an agreement or until one of them opts out of the negotiations. She concentrates specifically on the abilities of intelligent agents to coordinate their activities with other agents and to cooperate with them. These agents are assumed to be self-interested, rational, and autonomous. They do not share a common goal and each has its own set of preferences and acts according to them. Because of space constraints, only the first four chapters will be reviewed here.

In the introduction to the book, the author reviews some of the basic concepts from game theory, such as the extensive and strategic representations of a game. The coalitional representation is left to the references. Then, in chapter 2, she discusses in some detail the Rubenstein model of alternating offers, wherein there are N agents the need to reach an agreement on a particular issue, and the agents can only take actions at certain times determined in advance and known to the agents. There is no decision-regret, and the agents are provided with utility functions. The goal is then to find simple strategies that could be recommended to all agents so that no agent could benefit by an alternate strategy.

The author turns to negotiations about data allocation in chapter 3, where servers are autonomous and self-interested, can share documents, and need to make decisions where to locate data available to them. She assumes that the agents prefer any agreement in a given time period over the continuation of the negotiation process indefinitely, assumes losses of unused information, and only considers utility functions with fixed losses per unit time and a discount rate constant in time. The negotiation process is considered in cases of complete information, in which the servers know the expected usage of each dataset, but not the future usage, and in cases of incomplete information, where the expected usage is not, but only the past usage. Recognizing that the allocation problem is NP-complete in these cases, she brings in various techniques from optimization theory and artificial intelligence to deal with it. Simulations using hill-climbing, are shown to give better results than backtracking or genetic algorithms. She also gives a literature survey on distributed file allocation and with various other techniques for the incomplete information case.

In chapter 4, the author discusses bilateral negotiations for resources. In this scenario, one agent has access to a resource and is using it during the negotiation process, while another agent is waiting to use the resource. Cases of both complete and incomplete information are treated, along with cases of multiple encounters. The case of resource allocation is different than that of data allocation in that in resource allocation each agent always prefers a larger portion of the resource. Also, one of the agents loses over time while the other gains in resource allocation. The author discusses an interesting example dealing with the sharing of resources between NASA and ESA, and which illustrates the theorem that an agreement will be reached in the first or second period. This result is also different from the data allocation case, where in the latter, agreement is always reached in the first time period. The case of incomplete information is studied using the notion of "sequential equilibrium." This requires that an agent's strategy in each time period will be optimal given its opponents' strategies and its beliefs, and the history up the given time period. Three conditions are imposed on the sequence of strategies and the agent's system of beliefs that serve to characterize sequential equilibrium. The author again compares this with the data allocation, and concludes that for resource allocation with incomplete information, there is less incentive for telling the truth. The author then addresses the situation where the agents can meet several times in order to carry out the negotiation. She points out the use of "pooling" and "separating" equilibria in analyzing the situations of multiple encounters. An agent can have different utility functions, giving the agents different "types". If all these types select the same strategy in all states, this is a pooling equilibrium. If it is not, it is a separating equilibrium, and then it is possible to identify the agent's type from its actions. It is shown, as expected, that the negotiations in the multiple encounter case end no later than the second time period. Here again, the author uses the NASA and ESA robot examples to illustrate the results she derives for multiple encounters. She also shows the results of simulations to validate the agent's performance in situations of multiple encounters. Brief discussion is devoted to extensions of this model, including two agents with more than two encounters, multiple resources, and cases where there are more than two types of agents. For the case of many resources, the author concludes that the agent holding the resource will not stop negotiating with one agent and initiate a new negotiation process with another agent for a different resource. Other approaches to the resource allocation problem follow.

Synchronization is everywhere! This is the feeling one may get once alerted for it. Everyone is familiar with all kinds of biological rhythms ('biological clocks') that create some kind of conformity in time and in nature. This includes for instance neural activity and brain activity, but also the cardiac vascular system. Clearly, there are numerous other examples to be mentioned, sometimes much more controversial like the claimed synchronicity of the monthly period of nuns in a cloister, and so on.
Synchronous motion was probably first reported by Huygens (1673), where he describes an experiment of two (marine) pendulum clocks hanging on a light weighted beam, and which exhibit (anti-)frequency synchronization after a short period of time. Synchronized sound in nearby organ tubes was reported by Rayleigh in 1877, who observed similar effects for two electrically or mechanically connected tuning forks. In the last century synchronization received a lot of attention in the Russian scientific community since it was observed in balanced and rotors and vibro-exciters. Perhaps an enlightening potential new application for coordinated motion is in the use of hundreds of piezo-actuators in order to obtain a desired motion of a large/heavy mechanical set-up like for instance an airplane or Mm-scanner, or the coordination of microactuators for manipulation at very small scales.
In astronomy synchronization theory is used to explain the motion of celestial bodies, such as orbits and planetary resonances, In biology, biochemistry and medicine many systems can be modelled as oscillatory or vibratory systems and those systems show a tendency towards synchronous behavior. Among evidences of synchronous behavior in the natural world, one can consider the chorusing of crickets, synchronous flash light in a group of fire-flies, and the metabolic synchronicity in yeast cell suspension.
The subject of synchronization has received huge attention in the last decades, in particular by biologists and physicists. This attention probably centers around one of the fundamental issues in science, namely curiosity: how come we find synchronous motion in a large ensemble of identical systems? Also, new avenues of potential use of synchronicity are now being explored.
Synchronization has much in common - and is in sense equivalent to-coordination and cooperation. In ancient times it was already understood that joint activity may enable to carry out tasks that are undoable for an individual.
The authors' interest in the subject of synchronization is strongly influenced by a desire to understand what the basic ingredients are when coordinated motion is required in an engineering system. We therefore have concentrated in this book on synchronization or coordination of mechanical systems, like in robotic systems. This allows to delve, on the one hand, in the theoretic foundations of synchronous motion, but, on the other hand, made it possible to combine the theoretical findings with experimental verification in our research laboratorium.
This book concentrates therefore on controlled synchronization of mechanical systems that are used in industry. In particular the book deals with robotic systems, which nowadays are common and important systems in production processes. However, the general ideas developed here can be extended to more general mechanical systems, such as mobile robots, ships, motors, microactuators, balanced and unbalanced rotors, vibro-exciters.
The book is organized as follows:
Chapter 1 gives a general introduction about synchronization, its definition and the different types of synchronization.
Chapter 2 presents some basic material and results on which the book is based. In Section 2.1 some mathematical tools and stability concepts used throughout the book are presented. The dynamic models of rigid and flexible joint robots are introduced in Section 2.2, including their most important properties. The experimental set-up that will be used in later chapters is introduced in Section 2.3, where a brief description of the robots and their dynamic models is presented.
Chapter 3 addresses the problem of external synchronization of rigid joint robots. The synchronization scheme formed by a feedback controller and model based observers is presented and a stability proof is developed.
Simulation and experimental results on one degree of freedom systems are included to show the applicability and performance of the proposed controller. The main contribution of this chapter is a gain tuning procedure that ensures synchronization of the interconnected robot systems.
The case of external synchronization for flexible joint robots is addressed in Chapter 4. The chapter starts by explaining the differences between rigid and flexible joint robots and the effects on the design of the synchronization scheme. The synchronization scheme for flexible joint robots and stability analysis is presented. The chapter includes a gain tuning procedure that guarantees synchronization of the interconnected robot systems. Simulation results on one degree of freedom systems are included to show the viability of the controller.
The problem of internal (mutual) synchronization of rigid robots is treated in Chapter 5. This chapter presents a general synchronization scheme for the case of mutual synchronization of rigid robots. The chapter includes a general procedure to choose the interconnections between the robots to guarantee synchronization of the multi-composed robot system. Simulation and experimental results on one degree of freedom systems are included to show the properties of the controller.
Chapter 6 presents a simulation and experimental study using two rigid robot manipulators and shows the applicability and performance of the synchronization schemes for rigid joint robots. Particular attention is given to practical problems that can be encountered at the moment of implementing the proposed synchronization schemes. The robots in the experimental setup have four degrees of freedom, such that the complexity in the implementation is higher than in the simulations and experiments included in Chapters 3 and 5.
Further extensions of the synchronization schemes designed here are discussed in Chapter 7. Some conclusions related to synchronization in general and robot synchronization in particular are presented in Chapter 8.

This book provides a concise and self-contained introduction to post-modern control theory and estimation priniciples without burdening the lay reader with too much math.

The first section introduces system norms for multivariable control and the structured singular value for robust control design. The next section covers parameter estimation techniques. The material from both sections is then combined for application to practical design problems. All sections contain case studies of example design problems which are presented in detail.

The introduction to system norms is perhaps the best amongst existing books in terms of access to readers with only an undergraduate level background in mathematics. The section on system ID however might require parallel readings from other books if included in a course.

Overall, the book is perfectly taylored to introduce post-modern control philosophies to both graduate students as well as practising engineers.

This book provides a good introduction to the power of MATLAB to analyze control systems. It is directed mostly at the 400 level undergraduate, but the practicing engineer might find it of use as an introduction. It is a good balance between teaching the discipline of control systems and the uncovering syntax of MATLAB. MATLAB is pretty straight forward and and almost intuitive (except for its horrible printing commands). This book takes advantage of the fairly easy syntax to get on with some interesting engineering applications. As the title suggests, the applications are in the area of control systems asa encountered by the electrical engineer.

Wearable robotics is gradually being weaned from the exclusive confine of science fiction, to have a presence as an engineering field. Here, the text confines itself to the idea of biomechantronic exoskeletons. Take the instance of a bipedal walking robot. Its gait mimics deliberately that of a human. As the text says, this will require improvements in several areas. Like the micromachining of inertial sensors, to give the robot more information about its accelerations in real time. Plus, lighter motors and actuators are needed. And the motors should be more efficient, since the weight of a power supply greatly hinders the robot's usability and range.

Modelling of the human arm and hand is also surveyed. Along with the study of how humans do load bearing on their frames while moving. The intent is to copy as much of this as possible to a robot.

Ordered this with two other books, and i never recieved. Tracking info just says i should have recieved 2 weeks ago. How do you call Amazon???

This book is by far the most confusing, rambling, badly-written, and unedited programming manual I have ever seen. The information is just thrown together haphazardly and without logic. The author jumps randomly between subjects and there are no readable threads. Finding anything in this book is impossible. I took forty minutes trying to find out how to declare a variable, and finally threw the book in the trash. Don't waste money on it, period!

Lots of basic information as well as advanced techniques. If you are just getting into PIC's, this is a great text to start with. If you are already PIC fluent, it makes a great desk reference. The free PCB is a bonus!

This book is a very strong learning tool. An in depth study of Picmicro microcontrollers and how to program them. This is a must have book for anyone who wants to understand what goes on under the hood.

I bought this book without hesitation because I'd had contact with the author in the past, had other books by him, etc.,, The book depends on the CD, but the CD is not there, and can't even be downloaded as far as I can tell.

Myke says he's proud that they've included the PCB - why? Does anybody really want to gather up the parts to build this thing? The parts kits is not that great a deal either, if you can find it. Give me the **** CD which I need, not some stupid PCB I don't need - I can easily come up with a programmer, but I have no way to create the CD! In all the years I have been buying technical books, this is the stupidest move I've ever seen.

This book is a complete waste of money. Run away and boycott the publisher until they correct this rip-off.

From Myke Predko's web site:
(so where is it?)

"The CD-ROM that comes with this book is designed to be an integral part of the book. In the introduction I suggest that the reader load the CD-ROM into their computer before starting to read through the text as there is some unique information as well as code that is not present in the book due to space concerns. Along with the source (and executable) code for all the applications presented in the book, the CD-ROM also contains:

* HTML Interface to the contents of the CD-ROM including a page for each experiment and application
* HTML "appendices" for code "snippets" along with 16 Bit arithmetic and interfacing code examples and macros that can be "cut and "pasted" into your own applications.
* Two pdf appendices, "Introduction to Electronics" and "Introduction to Programming" for new developers
* Microchip MPLAB version 5.11, UMPS demonstration version 1.76 and gpsim/gpasm to give people different options in developing their own PICmicro MCU applications
* pdf data sheets for all the PICmicro MCU parts used in the book

The two pdf appendices, "Introduction to Electronics" and "Introduction to Programming" were written to help someone new to microcontrollers gain the background necessary to understand the concepts presented in the book. These appendices (which total over 250 pages) are virtually a stand-alone book on their own and provide reference information that experienced developers will probably find useful. "Introduction to Electronics" starts with basic electrical theory and explain the concepts behind digital logic, Analog to Digital Conversion, low-current power supplies, prototyping and basic test equipment. "Introduction to Programming" explains basic programming concepts and goes on to discuss structured programming and provides references to the "BASIC" (including PICmicro MCU varients) and "C" languages. "

I read this book from cover to cover, not because it was required by the Formal Languages course that I took, but because it is a very good read. This book gives all the necessary details in every theorem that it proves, which can be considered both a good and a bad thing, depending on your level of knowledge of the subject. I personally believe it is a good thing, because after reading any proof in this book you do not feel skeptical if it works or not, like it can be after reading a proof that skips lots of steps.

I haven't used their online resources, and I didn't do many problems from this book, because the professor teaching the course came up with problems of his own. However, from what I've seen, they have a very reasonable collection of problems suited for self-study. Every well established field has a list of standard problem, and Language Theory is no exception. The problems in this book certainly cover most of the standard ones. Please, also be aware that although this book is a good read, it is not necessary an easy read - be prepared to invest considerable amount of time into this book.

I cannot give this book 5 stars simply because I do not think it is much better than previous editions. As a matter of fact, I think it is worse. I did not have a very close look at previous editions, but I know for a fact that they were more rigorous and formal and covered more topics. As a result of that, they were less suited for teaching an introductory course, but some of the topics they studied there are really nice. From what I've heard about previous editions though, it seems that they described several open problems, that are no longer open. So I'd suggest getting this new edition, simply because it has more contemporary information.

Some people write in their review that this book requires solid background in the area it covers. I respectfully disagree - I had little to none background in Language Theory and Complexity Theory prior to taking this course and (consequently) starting to read this book; however, I did very well in the course and enjoyed it very much. Of course, I was lucky to have an excellent professor teaching that course. If good books came with good professors that would be a killer package, but unfortunately they don't.

I have not read any other books in this field, so I have nothing to compare with, but all in all, for me it worked great and if you have a good professor and are passionate about the subject, I'm sure this book won't be a miss.

This was used as a 3rd/4th year computer theory course at my university. For the most part the book was only used for homework problems where we got help from the TA and the professor had detailed slides. But from the times I tried reading the book to understand some problems it was literally like trying to read something in a different language. This book makes no attempt to make things understandable.

The previous edition of this text was published in the late 70's (1979), and it was still in use today in many schools and Universities across the world. For good reason too, the authors of this text really nail down the concept of computability as we understand it today. It is very difficult to find an undergraduate curriculum that does not include a course in Computability or theory of computation, and that is certainly a change from a couple of decades ago where this type of study was left to the Graduate level curricula. What this means to the reader is that one can not be a Computer Scientist without understanding the concepts and theory behind what computability really means.

Things like Context Free languages and grammar are used readily in things like XML and its accompanying standards such as the DTD. So, it makes sense to update a classic text to include such topics and further illustrate to the reader that what once was a theory is now center stage of Computer Science and the IT industry as a whole.

The text starts with the classics such as an introduction to automata theory followed by languages. The authors have taken a more relaxed approach to the topics as the proofs are less formal and easier to follow. Plain text is usually used to informally proof the topic at hand, and the authors go into a more formal approach on selected proofs. This is definitely a better approach than the other texts in the same topic that proofs are center stage of the discussion and the reader gets lost early on in the process. The text is easy to read for students, and easy to explain for the instructors. I remember when I took theory of Computation for my graduate work proofs were so convoluted and difficult to read that I had to spend many of nights trying to understand what the instructor was talking about in the class.

As one would expect, the book then goes into Turning Theory and Machine with the concept to computability and complexity. Well, the good news is that the authors' approach to the topic does not change; lots of explaining of the basics followed by a more detailed formal approach to the topic. All I need to say is that I wish my text was this reader friendly! Chapter 8, Introduction to Turing Machines, sets the ground work for the rest of the text. It explains reducibility and more importantly how to reduce a problem, something I have never seen in any other text in such detail! Automata and its relation to Turing Machine is depicted in detail, so there is no gap between the topics. What is interesting is that the authors close the loop with actually talking about, for example the Halting problem, in the real world with a program.

As one would expect, different classes of problems are explored in detail with many examples (theory and real-world examples) that accompany the topic at hand. Each chapter ends with a summary of topics discussed followed by a set of exercises. There are also a number of exercises at the end of each section in a given chapter in order to reel-in the topic for the reader.

All and all, this is one great text on automata and computation theory. It is easy to read and follow for the students without the loss of content. The authors relate abstract concepts to real-world examples to further illustrate the importance of the topic at hand.

This book is a one stop solution to your theoretical computer science needs (at least, as an introduction). If you're interested in language theory, deterministic / non deterministic finite state automata design, grammars and regular languages, computational complexity (temporal and spatial complexity), this the book for you. The formal notation used in the book is not the heaviest ever seen for this kind of subject, so it remains comprehensible (assumed it's not your first exposition to this discipline). I found it particularly interesting starting from chapter 8, when it covers turing machines, indecidibility in chapter 9 and intractability in chapter 10.
All in all, it's a good introduction to these concepts. I give it 4 stars because some proofs could have been easier, but this is not a big problem. The P and NP classes of problems are wonderfully explained. We are speaking about a book every computer scientist out there should have on his/her shelf. Those who consider this book extremely hard and difficult is because of their lack of fundamental knowledge in computer science. Of course, this is not the first book you should read on the subject. But be assured, this book will give you what it promises: a good knowledge about languages theory, indecidibility and intractability of problems.

I've heard that the first edition of this book is a classic. Reading the second edition, I can kind of see that -- occasionally there will be a stretch of 5 pages or so that is wonderfully clear, concise, and informative.

But overall, this edition is a disappointment. The explanations tend to be mechanical and unhelpful, and are sometimes confused or just incorrect. New sections on mathematical foundations and applications have been added, but there isn't really adequate space devoted to covering either topic, and the results are so rushed and lacking in context that I can't see those sections being useful to anyone who would need them in the first place. Finally, this edition needs to be proofread for correctness! It contains numerous mistakes, some of them in the presentations of key proofs.

When I bought the previous edition of this book, I couldn't put it down. I had no idea how easy PIC BASIC was, and that you can get PIC chips for free!

Every project works and there is a thorough explanation in plain english as to how it works. The projects covered in this book are really very cool and I highly recommend it to anyone just getting into this field.

This book is ok if want to build the robots explained in the book or useit to expand your knowledge. The reason I like this book is because he use's both the PICBASIC and PICBASIC PRO to program the robots. But it has too many errors and some of the projects are not fun enough to excite robot hobbyist at all. I wish he make another book which has more projects. Still the book is great!!!

This is a great book for learning a "beginner's basics" in microcollers. I give it four stars because of the misleading cover images of robots. There is nothing in the book about making a complete robot, but there are good basic microcontroller applications that could be used in a robot. In other words, there are PIC Basic programs where the reader can see real world applications while learning how to program and set up circuits.

I'm afraid I need to disagree with the professor's review. If you are at all interested in the PIC microcontrollers and use of PIC Basic you would be best served looking elsewhere for inspiration. John Iovine's book is poorly written in so many ways that even someone new to microcontrollers would find very little use for it. Aside from his writing style being somewhat confusing, this book is full of outdated information. With a publishing date of 2004 I was suprised to find so much information applied to applications and hardware from the 90's inside. Aside from this, which in and of itself is reason enough not to waste your time or money on it, he doesn't list any resources in his text. For example, he makes mention of serial LCD's, but doesn't provide any specific examples of vendors or models and makes claim that they all work the same, which is not true. His projects are the same basic beginners projects you can find all over the internet for free and are less descriptive than those you might find elsewhere. There isn't a single PCB design in the book and he demonstrates all of his projects with a breadboard. Most unfortunate, is the pictures he provides of the finished breadboards - they are taken at an angle, far enough away from the breadboard that not only could you not use the picture to validate your own breadboard design, you can barely tell what is what on the breadboard at all. On top of everything else, he explores only the PIC16F84 chip, which while one of the more popular chipsets ever produced by Microchip, is also outdated information since Microchip has since updated this to the PIC16F84A model and he doesn't cover any other chipset in any detail. One or two of his projects and examples show a different chipset in the schematic, but that's about it. His exploration of the BASIC language is no more or less descriptive than, say, the users guide of the PIC Basic language itself (which you can get for free). There are also many omissions and errors throughout the book, specifically as they apply to his projects. For example, his H-Bridge DC motor controller design is not smokeless (you could easily fry the PIC and/or other components in this design) and is missing several key components (like capacitors) that would prevent the design from working in a real world application. He also has a half a dozen or so shameless plugs throughout the book pointing you to 'other books he has written' to get more information - if this book is any example of what you might find in his other books, you'd be best to stay away from this author altogether. The cover of this book claims it is 'Completley updated and revised' - I guess this means the first edition was really horrible, the second edition is, in my opinion, simply not worth the read.

I bought the original edition of this book about 5 years ago, and it was instrumental in getting me started in PIC microcontroller programming and implementation. Although there are some errors throughout the book (someone commented that there are also errors in the new edition), and the hardware referenced is fairly dated (SPO256, for example, and some would argue the outdatedness of the PIC16F84), it's still a valuable source of information and reference. Browsing through the second edition, I see he's added PIC Basic "Pro" material, and filled out the chapters with more material specific to the chapter. The original edition was a good "hand holding" introductory tour through the use and function of a basic PIC microcontroller, and although I don't own the second edition yet, it looks to be more of the same, with a repeat of most of the original material.

The style of writer for explaining the contents is difficult to understand.
I need to repeatedly open the pages already read to understand the meaning of the contents.

But information of the resources and web info of the micom vendor and supplier is very usefull for me.

Si Ud. ya tiene experiencia con microcontroladores, este libro es de mucha uitilidad para revisar soluciones prácticas a problemas típicos. Yo personalmente ya realicé el experimento del control remoto (pág 371) y funciona muy bien, me estoy divirtiendo bastante y mi próximo objetivo es el NTSC.

A wealth of information! Such incredible projects! Video, Remote Control, Robotics, Servo, RS-232 Serial I/O. The author does a superb job at explaining many complex topics, and makes the instruction set easier with diagrams. The book is not really about the 8051 Microcontroller, but rather the family of 8051 compatible devices, which uses the Industry Standard 8051 set of instructions.

First off, a book should never be written in the first person. "I wrote this program...", "I did this better..", etc., etc. This guy loves to blow his own horn. More "I"s in this book than in a dark alley peep show. Very annoying to read with all of the "I"s and other bad grammar.

Second, if someone is new to the 8051 architecture this book is not for them. The author does a bad job of explaining many simple concepts, especially the 8051 memory layout. His diagrams and explanations are illogical and down right confusing. And when it came to more advanced topics such as serial IO via interrupts, hardly a peep of good information or examples.

There are many good books on 8051/8052 micro-controllers and this is definitely not one of them.

The author give me an impression that he just started to learn about 8051 right before starting to write this book.
In this book, it is not much explanation on the hardware features. Being able to connect to external memory is one of the biggest strength of 8051. Surprisingly, the author failed to eleborate on it. His book on PIC is a great book. However, I think he treats 8051 as PIC, which is not the objective of the 8051 book. The projects included in this book are more like PIC project, where small memory, I/O circuit is needed.

However, the programming portion in this book gives rather simple and not bad explanation to assembly language beginner like me. I think that is the only useful part in this book.

Hans Moravec 3D mapping technology will give computer depth perception; the capability of identify objects; and the ability to recognize texture, color, and material composition. Moravec initial 3D map prototype constructed a 256x256x64 cell volume equating to 4 million plus cells; the prototype used three cameras, produced a stereoscopic range, and generated 5,000 evidence rays in 5 seconds. The hope was the 3D map would allow robots to navigate effectively. CPU power was needed. 1,000 MIP is the minimum powered required to create the 3D map; the robot speed is a slow travel at this computational threshold; the 3D map will allow the robot to find doors, stairs, walls, and other 3D objects. 3D maps robots will be used by industrial companies and these robots will take the form of Automatic Ground Vehicles and fork-lift trucks and simple consumer vacuum cleaners.

Moravec three rules of robot success are: 1) The robot must be reasonable priced 2) The customer should not have to call in specialist to put a robot to work or to change its routine 3) The robot must be reliable for at least six months before encountering a problem or a situation requiring downtime for reprogramming or other alterations.

In the 90s, Dean Pomerleau built ALVINN, a neural network with 5,000 adjustable connections, whose desired desire was built to imitate a human driver; the NN output determined the steering position; some of the camera pictures simulated being further left in the lane with corresponding adjustments in steering; NN time to learn new roads was reduced to 5 minutes; the system provided neural interconnection weights for many road types. A new road type was determined by comparing the lower half of the image with the upper half and if they matched the road up ahead was the same locally, otherwise the new road type was added to the library. The NN input was a low resolution of the road using the blue from green substitution. In 1991, ALVINN traversed a busy 30km highway at 70km/h and Pomerleau earned his PHd.

Todd Jochem, a student of Pomerleau, built the next generation of code called RALPH. RALPH used 32x32 pixel low resolution picture of the road. The land ahead appeared as a wedge in the distance. If the road angles left or right, it estimated the blur in brightness changes, one cell from the next and the sharpest vector was kept. RALPH learning was instantaneous and driving became a technique of sliding over memorized vectors. RALPH drove from Washington DC to San Diego, 98.2% of the time in control, at an average speed of 100km/hr.

Rod Brooks declared the model based approach to robotics was unworkable. Brooks designed behavior control through layers, he called reflexes, for example, one behavior might cause the robot to steer away from an obstacle and another keep the robot traveling along the wall. The limiting ability to reflexive modeling was a limitation in cognitive ability, like a moth trapped in a street light. Brooks designed Cog which represented a larger number of learning reflexes allowing the robot to learn visually by imitation. Moravec thinks that reflexive technology will accomplish its desired goal, however, states, "I think there is a faster route, on that imitates at a higher level of abstraction" referencing conditioning modeling. Moravec concludes most practical automatic machines are behavior based.

The retina modeling is the benchmark breakthrough for the beginning of modern robotics. The retina is a centimeter across and a half millimeter thick and has 100 million neurons: horizontal cells which are light sensitive, narrower bipolar cells connected by Amacrine cells, and ganglion cell, which bundle to form the optic nerve. A million ganglion cells measure light intensity and differences over space and time. A 1000 MIPs machine could match the 10 scans a second.

1st generation robots will emerge around 2010 and possess 3,000 mips computation power; their size, shape, and strength will be human like; they will be efficient mobile devices on flat ground and able to traverse stair and manipulate everyday objects; and 2,4, and 6 legged robots will be able to cross most terrains and carry their own power supply, moving slowly, and for short distances. The robots will be heavy with perhaps three motors per limb. Movement may be done through shape bending alloys. A "Shape Bending alloy" bends at room temperature, but when heat is applied, it will return to its original shape with force. Robots will be able to perceive their surroundings with sensors, video camera configured for stereoscopic vision necessary to construct a 3D map and from the map it will be able to recognize locations, plan trajectories, and detect objects by color, shape, and location.

2nd generation robots will emerge around 2020 and have 100,000 MIPS, a 30 fold increase in computation power. 2nd gen robots will be capable of adaptive learning; the robots will adjust its behavior in response to the action past effectiveness, as the robot actual behavior is nudged closer to the human ideal. Robots will be packaged with learning models and probably be capable of being trained by humans through conditioning modules and these conditioning modules watch for desirable and undesirable situations that act on task oriented programs. Conditioning signals come in two categories: positive which raise the probabilities and negative which lowers the probabilities; character is a product of the suite of condition modules of he host. 2nd gen robots will be able to learned from 1st gen robots. 2nd gen robots will use central computer stimulations of robots, in action, to approximate results by gathering data and generalizing from the data, of other robots. A proper simulation would the result of thousands of learned models for various basic interactions and these simulations would be used to effectively construct condition suite by super central computers. 2nd generation robots will find jobs everywhere.

3rd generation robots will have 3 million MIPS and they will learn by faster through trial and error simulation, done by human supervision and super computers at the factory which will be capable of stimulating in real time. The robot will be able to recognize objects for what it is, so the proper interaction modules can be brought up called perception modules. Because these robots will be processing faster than real time they can run prediction simulations to determine if a response will turn out badly and alter its plan of action. In the spare time the computer could preplay previous experiences and try variations on them, learning new ways to improve performance and invent its own simple programs in response to a specialized conditioning module. Adaption is a process of corrective sequences of robot actions and how close they are to the desired end, very similar to the affects of genetic algorithms. These robots will need time to play and use their ability to adapt, imitate, and create simple programs of its own. They will have a theorm prover to find an absolutely correct solution, of arbitrary generality, subtlety, and deviousness, if one exists.

The 4th generation machine will have 100 million MIPS and advanced mechanized reasoning. These robots will write their own programs, understand natural languages, and understand concept and statements more deeply.

Philosophically Moravec wrestles with the word "IS". What is the purpose of this life? Mans purpose is too be born, learn good over evil, and gain increase through a family. Man environment provides beauty and enjoyment for man. A machine should never have dominion over a man. Moravec explains the purpose of man within the context of natural laws. He calls the natural laws stable, measurable, definable, and reliable. Any rationale beyond natural law is considered obscured. Existence cannot be explained by natural laws only. Even Moravec cannot advocate annihilation and clings to the idea that his consciousness will continue either in another form or through robots. God is the reason for mans existence and man exists to become like God. Since God exists than natural laws must be lower level laws. Moravec theorm is incomplete considering the final destination of man.

Man exists to chose between Good from Evil. A conditioning robot cannot expect to achieve this discernment unless higher moral laws govern it. The acquisition of intelligence is beneficial within a natural law sphere but does not necessary suggest the robot will be capable of choosing good over evil. The devil is very intelligent, yet he did not chose good over evil. If a man is more intelligent than his parents, do we call him better?

Suppose, a mans interactions are evil but the results are good, do we call him justified? If robots convert all matter into digital virtual reality, do we say ou existence has improved? An existence that is force upon us. Intelligence must yield to agency which is the freedom to act and not be acted upon. Intelligence alone can not to the reason for existing, intelligence is only part of the meaning of existence, choice and accountability is the larger portion of existence. Man choice is to learn and to discover the "why and how" knowledge necessary, too reject evil. This is not an automated task which can be programmed because opposition and temptation complicate the algorithm into a chaotic mess, of uncertain and solid morality, for an hedonist. A robot will not know how to choice good and evil because it can know sense a higher purpose and morality, so its action will not follow a higher purpose.

The best book on the future of robotics and automation! However, Hans believes robots are our wonderful mind children and should grow into powerful machines that evolve quickly past us. He is then horrified that some humans may transform themselves into machines and become very dangerous. Why won't his mind children be just as dangerous or more dangerous? At least a mind-transferred human might seek pleasure and fun. While Hans' logical AI robots make their galactic invasion plans!

Why not engineer automation to its pleasure giving limits? Instead of giving robots a high quality of life, design automation to increase EVERYONE quality of life and wealth on Earth???

We all wish to know what will ultimately become of us, of that which we care about, the people we love.
One way Mankind has of receiving answers to this is through Religion.
Another way is through speculating on the basis of scientific knowledge and understanding.
Here the Speculations are preceded by a survey of the current state of Robotics.
This is preliminary to a set of projections of the distant future in which biologically- based beings i.e. us , are going to be not supplemented but essentially transcended and replaced by silicon- system artificial intelligences, robots of Intelligence far beyond our own.
The old- style humans , those who choose not to somehow transmit their identities into the new ' super- silicon beings' will kind of hang on as patronized parasites enjoying life as one big freebie thanks to their successful successors.
At this point some of us ' cool' to what is to come.
Magnificent minds simulating scenarios of infinite alternative lives simply do not warm our old aging hearts.
The prospect of monstrously beautiful recombinations in hyperspace of cyberbeings just does not turn us on.
Our minds are in the more mundane, the smaller seeings of our own inner poetries, the lives we make the people we love.
This kind of speculative stuff seems a minor curiosity when measured against the thick, dense , impossibly , non- controllable unpredictability of our small everyday lives.
Forgive us, Future- see-ers of the great Machine- meaning, we are staying home with our own for now.

I'll readily and happily admit that I'm no expert in robotics or the theory of Artificial Intelligence; I've had exactly one course in the subject, and know most of what I know thanks to Scientific American, Popular Mechanics, and the like. With that caveat out of the way, I can say with absolute certainty that THIS BOOK IS RIGHT! THE ROBOTS ARE COMING! RUN FOR YOUR LIVES!

Whoops, sorry, that's just wishful thinking. Seriously, this is a good book, well-written and interesting throughout. Though I personally felt that Moravec got a bit spacey (pun intended, if you've read the book) towards the end, the possibilities he raises are fascinating. As to how temporally accurate his predictions are, again, I can't say, though robotic vacuums did arrive essentially on schedule, and in general most of what he suggests seems feasible.

Like some of the other reviewers, I appreciate a book that runs counter, in large part, to the 'end-of-humanity' theme that seems to accompany the idea of robots gaining mind. As cool as "The Matrix," "Terminator," or "I, Robot" might be on the screen, a real-life instantiation of those themes would be less than cool. Being a fan of Occam's razor myself, I don't know that I'd expect the robots to expend enormous amounts of energy to enslave or exterminate us, assuming we didn't make ourselves too much of a nuisance, an this seems to be the tune Moravec himself sings.

Anyhow, this is a book that is occasionally puissant, hardly ever dull, and often thought-provoking. Any potential buyers may want to wait a few years, though, to see if Moravec keeps on schedule and releases a new version, as per his established pattern.

With high praise from such giants as Sir Arthur C. Clarke and Doctor David Brin on the dust jacket, I asked myself where I, unlettered and relative to them barely conscious, think I'm going trying to write a review. I have a friend who likes to say he never lets ignorance stop him from expressing his opinion on a subject. Guess I remember that one `cause it fits me so well, so here goes.
In his 1950 paper Computing Machinery and Intelligence, Alan Turing grouped the arguments opposing the possibility of machine intelligence into the following nine categories:
1- The Theological Objection - thinking is a function of the soul. Machines have no souls, so cannot think.
2 - The "Heads in the Sand" Objection - Thinking machines cannot be possible because the consequences would be too dreadful.
3 - The Mathematical Objection - Mechanical reasoning has certain provable limitations that human thought may not share.
4 - The Argument from Consciousness - Machines have no inner experiences to give meaning to their utterances, actions, or internal operations.
5 - Arguments from Various Disabilities - Machines will never be kind, moral, joyous, perceptive, original, etc.
6 - Lady Lovelace's Objection - Computers do only what we program them to do.
7 - The Argument from Continuity in the Nervous System - Nerves respond to arbitrarily tiny signal differences, while computers work in fixed-size steps.
8 - The Argument from Informality of Behavior - It is not possible to specify for a machine what to do in every possible circumstance a human might encounter.
9 - The Argument from Extrasensory Perception - Humans sometimes sense remote or future information unavailable to deterministic processes in computers.

Moravec provides current arguments countering each item above, but central to all seems to be this: the principle difference between human and machine is we are conscious. This state, however, is so complex we are unable to explain it. Neither do we understand how or from where it arises in our brains.
The author offers a compelling posit; If as of Robot's publication (1999), the most powerful computers could process a million MIPS (million instructions per second), computers capable of a billion MIPS should be just over the horizon. It will be then, Moravec projects, that the mysterious and exclusively human state we call "consciousness" will be revealed to be not exclusive at all, but merely the capacity to accumulate, process, and interpret sufficient amounts of data in the span of each instant of time - and that when this is achieved, computers will sense the state of their surroundings and thus become "conscious" in the same way we are.
He lays the groundwork for this leap carefully, detailing his personal experiences in robotics and the pace of advances in the field. Arriving at the present day situation, he then takes us step by careful step into the future. It's all completely understandable and reasonable. He's right - know what I'm saying?
Eventually though, his vision of the future exceeds my ability to absorb. I confess to less than a complete understanding of his universe of the future. One thing I did get loud and clear: there were no humans there.
Consider robots an intellectual mutation. These creatures we make will first surpass and then replace us, become us, probably in very much the same way we ourselves replaced the less capable lifeforms we arose from in the distant past. It's not a grim future the author envisions for humanity; it's a comfortable even spiritual retirement. Refuse to accept this, and you'll need to deny Darwin's theories too. Think about it.

Art Tirrell is the author of the underwater adventure novel "The Secret Ever Keeps" which does not contain robotics but does contain "...Simply put, the best underwater scenes I've ever read..." Meg W, reviewer.