Independent or dependent?

Chapter Independent or dependent?

Teach Yourself Electricity and Electronics Third Edition Book
Pages 748
Views 14,504
Downloads : 46 times
PDF Size : 4.4 MiB

Summary of Contents

Teach Yourself Electricity and Electronics Third Edition Book

  • There are some situations in which autonomous robots don’t work well. In thesecases, many simple robots, all under the control of one central computer, can be used.They function like ants in an anthill or bees in a hive. The individual machines are stu-pid, but the group as a whole is intelligent.Fourth generation and beyondNakano did not write about anything past the third generation of robots. But wemight mention a fourth-generation robot: a machine of a sort yet to be deployed.An example is a fleet or population of robots that reproduce and evolve. Past that,we might say that a fifth-generation robot is something humans haven’t yetdreamed of—or, if someone has thought up such a thing, he or she has not publishedthe brainstorm.Independent or dependent?Years ago, roboticist and engineer Rodney Brooks became well known for his workwith insect robots. His approach to robotics and AI was at first considered unortho-dox. Before his time, engineers wanted robotic AI systems to mimic human thoughtprocesses. The machines were supposed to stand alone and be capable of operatingindependently of humans or other machines. But Brooks believed that insectlike in-telligence might be superior to humanlike intelligence in many applications. Supportfor his argument came from the fact that insect colonies are highly efficient, and of-ten they survive adversity better than supposedly higher life forms.Insect robotsInsect robots operate in large numbers under the control of a central AI system. Amobile insect robot has several legs, a set of wheels, or a track drive. The first ma-chines of this type, developed by Brooks, looked like beetles. They ranged in sizefrom more than a foot long to less than a millimeter across. Most significant is thefact that they worked collectively.Individual robots, each with its own controller, do not necessarily work well to-gether in a team. This shouldn’t be too surprising; people are the same way. Professionalsports teams have been assembled by buying the best players in the business, but theteam won’t win unless the players get along. Insects, in contrast, are stupid at the indi-vidual level. Ants and bees are like idiot robots, or, perhaps, like ideal soldiers. But ananthill or beehive, just like a well-trained military unit, is an efficient system, controlledby a powerful central brain.Rodney Brooks saw this fundamental difference between autonomous and collec-tive intelligence. He saw that most of his colleagues were trying to build autonomous ro-bots, perhaps because of the natural tendency for humans to envision robots ashumanoid. To Brooks, it was obvious that a major avenue of technology was being ne-glected. Thus he began designing robot colonies, each consisting of many units underthe control of a central AI system.Brooks envisioned microscopic insect robots that might live in your house, comingout at night to clean your floors and countertops. “Antibody robots” of even tinier pro-portions could be injected into a person infected with some previously incurable 646 Robotics and artificial intelligence