X hits on this document

373 views

0 shares

0 downloads

0 comments

68 / 136

68

  • ‚óŹ

    Exploratory Workshop on the Social Impacts of Robotics

require much more research and development.

sible

to

improve

the

mechanical

accuracy

of

It may be

robots,

pos- and to

improve

servo

performance

with

little

more

than

careful

en-

gineering.

But

much

more

fundamental

research

and

develop-

ment will modeling,

be required before the sensor,

software

generation

systems

contro1,

interface

internal and mobil-

ity problems are solved. technology to improve the

Much remains performance,

to be done in

reliability)

sensor

and

cost

effectiveness of all types of sensory more remains to be done in improving the

transducers. Even speed and sophisti-

cation of sensory processing algorithms and special

hardware for both in space

recognizing

and

time.

The

features

computing

and analyzing

power

that

purpose patterns is re-

quired for

high

terns will

even

speed processing of visual and acoustic pat- require new types of computer architecture.

Sensory-interactive various kinds of

control systems that can respond to sensory data at many different levels of

abstraction are still

Current

commercial

very much in the research phase. robot control systems do not even allow

real-time sory data.

six-axis incremental movements in None have convenient interfaces

response by which

to sen- sensory

data of many different kinds can be introduced into the vo loops on a millisecond time scale for true real-time

ser- sen-

sory tems

interaction.

None of the commercial robot control sys-

have

anything

approximating

CAD

data

bases

or

computer

graphics current

models of the environment and workplaces.

programming

techniques

are

time

consuming

Finally,

and

not

capable of dealing with internal knowledge or sensory interactions.

sophisticated

These

are

very

complex

problems

that

will

require

many

person-years

of research effort.

It is thus not surprising

that

the

robot

applications

are

still

extremely

limited,

WHAT LIES IN THE FUTURE? All of the problems listed above are amenable to

solution.

It is only a matter of time and expenditure of resources be-

fore duce

sensors and control dexterous, graceful,

systems are developed that can pro-

skilled

behavior

in

robots.

Even–

tually, robots will be able to store about the world that will enable them

and recall knowledge to behave intelligent-

ly and even to show

a measure of insight regarding the

spa-

tial

and

temporal

relationships

inherent

in

the

workplace.

High order languages,

phisticated

control

computer-aided-instruction,

systems

will

eventually

make

it

and so- possible

to instruct robots that one might use

using much in talking

the same vocabulary and to a skilled worker.

syntax

There is robotics

no question that given enough

will

eventually

become

a

time and

resources

significant

factor in

Document info
Document views373
Page views373
Page last viewedWed Dec 07 14:44:02 UTC 2016
Pages136
Paragraphs11095
Words48033

Comments