Intelligent Systems

boorishadamantAI and Robotics

Oct 29, 2013 (3 years and 7 months ago)

102 views

Intelligent Systems

Lecture 13

Intelligent robots

Classification of robots

Industrial

Classification of robots on
purpose


For research

Home and office

Military

Searching

-

welding

-

painting

-

loading
-


unloading

-

transport

-

assembling


-

For space

-

For cracks


-

For land
reconnaissance

-
aerial
reconnaissance

-

For land tactic
operations

-

For air tactic
operations

-

For space

-

For underwater
operations


-

robots
-
toys

-

For care of
children

-

For care of elders

-

robots
-
guards

-
multi
-
purpose

home robots


-
robots for
soccer

-

Battle
robots

-

For
research of
behavior and
cognition

-

For
research of
navigation
and planning


Classification of robots on
mechanics

Classification of robots


Mobile


S
tationary


Programmed

(
without AI
)


Learned

(
without AI
)


Learned

(
with AI
)


Self
-
learning (with
AI)

Functions of control system of robot


Perception and recognition of entities of
environment


Interaction with human


Planning and replanning of behavior


Navigation, control of goal
-
seeking
behavior


Control of engines (motors)


Learning, forming of model of environment


Interaction with other robots and
equipment

Kinds of learning


Supervised


Teacher show how system must to answer on input
data (what to do in any situation)


Unsupervised


System itself finds laws in data


Reinforcement


System selects behavior on base award obtained
from environment and estimation of state of
environment (on base on interaction with
environment)


Kinds of planning


Planning systems are problem
-
solving algorithms that operate on explicit
propositional (or first
-
order) representations of states and actions. These
representations make possible the derivation of effective heuristics and
the development of powerful and flexible algorithms for solving problems.


The STRIPS language describes actions in terms of their preconditions
and effects and describes the initial and goal states as conjunctions of
positive literals. The ADL language relaxes some of these constraints,
allowing disjunction, negation, and quantifiers.


State
-
space search can operate in the forward direction (
progression
) or
the backward direction (
regression
). Effective heuristics can be derived
by making a subgoal independence assumption and by various
relaxations of the planning problem.


Partial
-
order planning (POP) algorithms explore the space of plans
without committing to a totally ordered sequence of actions. They work
back from the goal, adding actions to the plan to achieve each subgoal.
They are particularly effective on problems amenable to a divide
-
and
-
conquer approach.

The agent
-
environment
interaction in reinforcement
learning




Features of reinforcement learning and
main concepts


Learning is combined with working


Working is a sequence of actions


Plan of actions is
policy



Plan (policy) may be corrected in every time (step)


Action

is selected from policy (or no) in according to
estimation of
state of environment

(or
estimation of
action

in same state) and
reward

received from
environment


Estimation of environment is determined by goal (
target
)


Estimation of environment or action is executed with
delay after obtaining of award

Definition of planning

Relationships among learning, planning, and
acting









Traditional (to 1985) decomposition of a mobile
robot control system into functional modules

Brooks: “The key idea from intelligence is:

Intelligence is determined by the dynamics of
interaction with the world.”

A decomposition of mobile robot control
system based on task achieving behavior

Principles formulated by Brooks (1991) for behavior
-
based
robots


There is no central model maintained of the world. All data is distributed over
many computational elements


There is no central locus of control


There is no separation into perceptual system, central system, and actuation
system. Pieces of the network may perform more than one of these functions.
More importantly, there is intimate intertwining of aspects of all three of them.



The behavioral competence of the system is improved by adding more behavior
-
specific network to the existing network. We call this process layering. This is a
simplistic and crude analogy to evolutionary development. As with evolution, at
every stage of the development the systems are tested
-
unlike evolution there is a
gentle debugging process available. Each of the layers is a behavior
-
producing
piece of network in its own right, although it may implicitly rely on presence of
earlier pieces of network.



There is no hierarchical arrangement, i.e., there is no notion of one process
calling on another as a subroutine. Rather the networks are designed so that
needed computations will simply be available on the appropriate input line when
needed. There is no explicit synchronization between a producer and a consumer
of messages. Message reception buffers can be overwritten by new messages
before the consumer has looked at the old one. It is not atypical for a message
producer to send 10 messages for every one that is examined by the receiver.



The layers, or behaviors, all run in parallel. There may need to be a conflict
resolution mechanism when different behaviors try to give different actuator
commands.



The world is often a good communication medium for processes, or behaviors,
within a single robot.


Tasks and features of humanoid robots


Being a mobile robot with power supply and computer
control on
-
board


Navigating and moving in an environment made for
humans


Biped walking in a humanoid style


Gripping and manipulating objects designed for humans


Cooperative working with humans


Interacting with humans without endangering their safety


Having autonomous behavior


Communicating with humans in a simple and intuitive
way


Using a stereo
-
vision system as main sensor system


Using learning and adaptive behavior strategies


Using human
-
like intelligence


Having a design pleasing to real humans

Architecture of control system

Functional structure of control system

Features of control systems of
Sony’s robots


Adaptive control of
movement in real time


Selection of kind of
gait (walk) in real time


Possibility of
perception of space of
real world in real time


Multi
-
modal
interaction with
human

Main behavior systems of dog,
investigated by Sony during development
of Aibo

Internal motivational variables

Modules within Defense
-
Escape mode


Modes comprising Agonistic Subsystem


Role of Drives in Behavior Selection


Connection of emotion with
behavior

Using of emotions in selection of behavior

Objects used in experiments:


Meat (red)


Water (blue)


Owner (green)

Selection of behavior

Tree of behaviors

Architecture of EGO

Storing and using of association
between visual image and its name

Architecture of humanoid robots of Sony

Recognition of multi
-
faces

Emotion
-
based behavior of robot
SDR
-
4X