Skip to main content icon/video/no-internet

Layered Control Architectures

A layered control architecture is one in which there are multiple levels of control at which the sensing apparatus is interfaced with the motor system. It is distinguished from hierarchical control by the constraint that the architecture should exhibit dissociations, such that the lower levels still operate, and exhibit some sort of behavioral competence, in the absence (through damage or removal) of the higher layers but not vice versa. A substantial body of the neuroscience literature can be interpreted as demonstrating layered control systems in the vertebrate brain; layering is also an important theme in the design of artificial control systems, for instance, for autonomous robots.

Layered Control in the Vertebrate Brain

In many ways, the notion of layering is a common, often unspoken, assumption in neuroscience; however, the implications of the layered nature of the brain are not always acknowledged in a field often dominated by the study of the mammalian cortex. The idea of the brain as a layered architecture can be traced to the views of John Hughlings Jackson, a 19th-century neurologist inspired by the Darwinian revolution. According to Jackson, the various anatomical levels of the brain implement multiple functional levels of sensorimotor competence with higher centers supporting the same sort of “sensorimotor coordinations” as those below, but in a more complex fashion. This view contrasted with that of many of Jackson's contemporaries for whom the highest levels of the brain, particularly the frontal lobes, were considered to be the seats of understanding and consciousness. Jackson strongly asserted that, although the frontal lobes may play a role in these more refined functions, the brain is a product of evolution and, therefore, all of its levels must be involved, in some way, with the coordination of sensing with action. Jackson also popularized the idea of dissociation; he argued that a breakdown at a higher layer should cause a reversion to the next highest layer of control. Discoveries of such dissociations between the vertical layers of the vertebrate nervous system were among the first findings of neuroscience. For instance, removing the cortex from a cat or a rat eliminates many major sensory, motor, and cognitive centers but leaves intact the ability to generate basic motivated behavior. That is, the animal still shivers when cold, escapes or fights when attacked, searches for food and drink, stops eating when sated, and so on. When most of the forebrain is removed, integrated behaviors can no longer be generated, but the capacity for elementary actions (walking, grooming, eating, etc.) is spared. With all but the hindbrain and spinal cord removed, the animal cannot coordinate the movements required for these actions; however, most of the component movements that make up the actions are still possible. The notion of a layered architecture is now being mapped out in more detail in the context of specific types of behavior. For instance, as illustrated in Figure 1, the vertebrate defense system—the control system that protects the body from physical harm—can be viewed as being instantiated in multiple layers from the spinal cord (reflexes), through the hindbrain (potentiated reflexes), midbrain (coordinated responses to species-specific stimuli), forebrain (coordinated responses to conditioned stimuli), and cortex (modification of responses according to context). In this system, the higher layers generally operate by modulating (suppressing, potentiating, or modifying) responses generated by the lower layers.

...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Sage Recommends

We found other relevant content for you on other Sage platforms.

Loading