Robot Brains

This book presents practical design guidelines for the creation of non-numeric, autonomous cognitive machines, examining in detail component parts, realization principles and providing real-world examples for designers, researchers and advanced students in the field. Haikonen starts with an introduction to the topic of cognitive machines before moving on to examine associative neural networks, including a look at the limitations of traditional neural associators. Following this, basic circuit assemblies are described, as a building block for systems introduced later in the book. The next chapter presents perceptory processes required to understand the cognitive system, leading to a discussion of motion perception, setting out how motors and moving parts can be interfaced with the associative neural system in order that fluent motion based on perceived affordances can be achieved without numeric computations. The final chapters focus on the increasingly complex issues related to cognitive machines, such as the concept of machine emotions and natural' language in robot brains. The last chapter explains how all of these component parts are related to conscious machines.