Sweet Basketball Song

Intelligent Robots Architecture "three layers"



Obviously, the complexity of the environment has a close relationship with the complexity of the control system that governs the behavior of the machine. Indeed, if the robot has to react quickly and intelligently in a dynamic and challenging environment, the problem becomes very difficult to control. If the robot, however, need not respond quickly, reducing the complexity required to develop the [Mataric, 2002].

The "three-tier architecture" can consider three different kinds of processes [Davis, 1998] [Sloman, 1996], similar to the neurological hypothesis MacLean triune brain or Freud's psychological hypothesis of the interaction between the "It" (or collection of impulses instinctive), the "Super-I" (which includes the objectives and social prohibitions high level) and "Yo" (which resolves conflicts or deleted them) [Minsky, 2000].

is likely that these processes operate in parallel and through many types of interaction. Are basically three (Moriello, 2005, p. 194 / 5):

· "Reactive": This is the most bajo y equivale a las partes más viejas y primitivas del cerebro humano (el paleoencéfalo o cerebro reptiliano). La información se adquiere por medio de los sensores externos e internos y se traduce en órdenes para los efectores. Se trata de procesos muy rápidos, o inmediatos, que permiten una respuesta directa a los eventos sensoriales.

· “Deliberativo”: Es el nivel intermedio y equivale a las partes un poco más modernas del cerebro humano (el mesencéfalo o cerebro mamífero). Se podría decir que es el verdadero corazón del sistema, ya que provee las capacidades de razonamiento, planificación y resolución de problemas. Son aquellos procesos que desempeñan un papel importante in what is known as "thinking" over other processes related to the management of low-level actions.

· "Reflective" is the highest level and is equivalent to the most recent human brain (the neocortex or brain neomamífero). Allows the monitoring, evaluation, and some control of internal processes and the provision of typically human capabilities (such as introspection, learning and explicit decisions to redirect the attention.) It's basically learning mechanisms, one way or another, permeates the entire architecture.