Characteristics and connotations of reactive behaviors

Characteristics and connotations of reactive behaviors

As seen earlier, a reactive robotic system decomposes functionality into behaviors,
which tightly couple perception to action without the use of intervening
abstract (global) representations. This is a broad, vague definition.
Over the years, the reactive paradigm has acquired several connotations and
characteristics from the way practitioners have used the paradigm.
The primary connotation of a reactive robotic system is that it executes
rapidly. The tight coupling of sensing and acting permits robots to operate
in real-time, moving at speeds of 1-2 cm per second. Behaviors can be
implemented directly in hardware as circuits, or with low computational
complexity algorithms (O(n)). This means that behaviors execute quickly regardless
of the processor. Behaviors execute not only fast in their own right,
they are particularly fast when compared to the execution times of Shakey
and the Stanford Cart. A secondary connotation is that reactive robotic systems
have no memory, limiting reactive behaviors to what biologists would
call pure stimulus-response reflexes. In practice, many behaviors exhibit a

fixed-action pattern type of response, where the behavior persists for a short
period of time without the direct presence of the stimulus. The main point is
that behaviors are controlled by what is happening in the world, duplicating
the spirit of innate releasing mechanisms, rather than by the program storing
and remembering what the robot did last. The examples in the chapter
emphasize this point.
The five characteristics of almost all architectures that follow the Reactive
Paradigm are:
1. Robots are situated agents operating in an ecological niche. As seen earlier in
Part I, situated agent means that the robot is SITUATED AGENT an integral part of theworld. A
robot has its own goals and intentions. When a robot acts, it changes the
world, and receives immediate feedback about the world through sensing.
What the robot senses affects its goals and how it attempts to meet
them, generating a new cycle of actions. Notice that situatedness is defined
by Neisser’s Action-Perception Cycle. Likewise, the goals of a robot,
the world it operates in, and how it can perceive the world form the ecological
niche of the robot. To emphasize this, many robotic researchers say
ECOLOGICAL ROBOTICS they are working on ecological robotics.
2. Behaviors serve as the basic building blocks for robotic actions, and the overall
behavior of the robot is emergent. Behaviors are independent, computational
entities and operate concurrently. The overall behavior is emergent: there
is no explicit “controller” module which determines what will be done, or
functions which call other functions. There may be a coordinated control
program in the schema of a behavior, but there is no external controller
of all behaviors for a task. As with animals, the “intelligence” of the robot
is in the eye of the beholder, rather than in a specific section of code.
Since the overall behavior of a reactive robot emerges from the way its
individual behaviors interact, the major differences between reactive architectures
is usually the specific mechanism for interaction. Recall from
Chapter 3 that these mechanisms include combination, suppression, and
cancellation.
3. Only local, behavior-specific sensing is permitted. The use of explicit abstract
representational knowledge in perceptual processing, even though it is
behavior-specific, is avoided. Any sensing which does require represenEGO-
CENTRIC tation is expressed in ego-centric (robot-centric) coordinates. For example,
consider obstacle avoidance. An ego-centric representation means that it
does not matter that an obstacle is in the world at coordinates (x,y,z), only

where it is relative to the robot. Sensor data, with the exception of GPS, is
inherently ego-centric (e.g., a range finder returns a distance to the nearest
object from the transducer), so this eliminates unnecessary processing to
create a world model, then extract the position of obstacles relative to the
robot.
4. These systems inherently follow good software design principles. The modularity
of these behaviors supports the decomposition of a task into component
behaviors. The behaviors are tested independently, and behaviors
may be assembled from primitive behaviors.
5. Animal models of behavior are often cited as a basis for these systems or a particular
behavior. Unlike in the early days of AI robotics, where there was a
conscious effort to not mimic biological intelligence, it is very acceptable
under the reactive paradigm to use animals as a motivation for a collection
of behaviors.