ABSTRACT
This chapter provides a comparison of intelligent and cognitive agent architectures at the functional and design level, as well as reviewing intelligent agent architecture advancement until the year 2016. The basis for an artificial agent to have human-like decision-making characteristics is due to an essential element that is known as agent architecture, the blueprint for constructing an agent having reasoning skills. Wooldridge, in 1995, discussed agent architecture as a software design that is anticipated to provide human-like rational decision-making processes The progress of the Internet and computer technologies has increased the need for artificial agent-based distributed, dynamic, coordinated, and heterogeneous intelligent systems. Reactive agent architecture is grounded on the straight mapping of perception states into action. Brooks’ basic idea of subsumption architecture is that agents can generate intelligence in performance without explicit knowledge illustrations and abstract inference by using symbolic artificial intelligence techniques.
