Page 14 - Designing Autonomous Mobile Robots : Inside the Mindo f an Intellegent Machine
P. 14

Foreword






               One of the most exciting challenges a designer can face is that of building a practical
               autonomous mobile robot. This is as close as we mere mortals can come to producing
               a living being. Make no mistake; we are talking about very primitive beings, but
               artificial beings nonetheless. Autonomous robots exist today that can perform
               complex tasks without human assistance for weeks or even months, but these robots
               will seem laughably crude in the years to come.

               The building blocks for autonomous robots have been readily available for several
               years. Powerful microprocessors, laser-based sensors (lidar), Ethernet radio commu-
               nications, video processors, and a host of other subsystems are now priced at levels
               that permit practical autonomous machines to be built for an exciting range of
               commercially viable applications. There are even a wide range of simple sensors and
               actuators that allow the hobbyist to develop small, but sophisticated robots.
               The challenge is in understanding how these systems can be made to play together
               in a coherent and effective way to create a system that is far more than the sum of its
               parts. If the designer thinks of a new robot design as being laser-guided, or as using
               GPS navigation, the result will be a design that is inflexible. Such a design may be
               useful, but it will be not able to grow beyond its initial concept. A stripe following
               “Automatic Guided Vehicle” is an excellent example of such a design. Autonomous
               robots are much more robust and interesting beasts.

               It is my experience that any good concept will have an intrinsic elegance. A good
               software and hardware structure is like a snowflake, with each subsystem having the
               same basic structure as every other subsystem. At the center, a few basic structures
               hold it all together. Each point of the snowflake will have differences from the
               others, but will follow the same basic pattern.





                                                        xiii
   9   10   11   12   13   14   15   16   17   18   19