Sharing Knowledge in the Intelligent Robot Life-cycle

(1)

Sharing Knowledge

in the Intelligent Robot

Life-cycle

Ricardo Sanz

WOSRA

Collaboratively Working towards Ontology-based Standards

for Robotics and Automation

(2)

Abstract

In the life-cycles of intelligent robots there are two

knowledge-based systems of maximal relevance: the

intelligent robot system

and the

engineering system

.

The intelligent robot uses knowledge in the performance

of its tasks and in its relations with other agents. The

(3)

Purpose

• My purpose (in this workshop) is to

describe some

facts

about the

engineering of intelligent robots

• These facts concern

knowledge

and the

use of

(4)

Content

• Intelligent robot systems

• Engineering robot systems

• Roles of knowledge

• Systems and Flows

• Challenges

(5)

(6)

Intelligent/autonomous/cognitive

• I use

intelligent

in the title to convey a focus on

knowledge

• I would have used “

cognitive

” if not a too-human

(7)

(8)

Intelligent Robot Systems

• Robots (and systems beyond robots) can use

knowledge

• In general we can speak of “

intelligent

systems

”

• Intelligent systems implement a

cognitive loop

in the performance of

their service: perceive-think-act

(9)

Engineering robot

systems

(10)

What are systems?

• … A

“

system

”

is a construct or a collection of different

elements that together

produce results not obtainable

by the elements alone

. The elements include

hardware,

software, facilities, people, policies and information

. The

results include system level

requirements

, properties,

(11)

Two Views of the Robot System

(12)

(13)

Two/Three (One + (One + One))

Systems

• In the life-cycles of intelligent robots there are

two

knowledge-based systems

of maximal relevance:

the

intelligent robot system

itself and the

engineering system

.

• This last one can be subdivided in the

engineering

team

and the

engineering Project

(14)

A System-of-Systems Endeavour

The Team

The Project

(15)

System

Life-Cycle

Exploration

Concept

Development

Production

Utilization

Support

(16)

Life-cycle

ISO/IEC 15288: “4.1.24

life cycle

-evolution of a system, product, service,

project or other human-made entity from

conception through retirement”

(17)

“Life-cycle” concept value

• The purpose in defining the system life cycle is to

establish a

framework for meeting the

stakeholders’ needs

in an orderly and efficient

manner by the engineering team.

• Usually done by defining

life‐cycle stages

and using

(18)

Stage Purpose Processes

Exploration Identify stakeholders’ needs Explore ideas and technologies

The detailed SE processes applied during the life cycle model stages can be tailored and expressed in terms of the ISO/IEC/IEEE 15288 processes and their outcomes,

relationships, and sequence.

Concept Refine stakeholders’ needs Explore feasible concepts Propose viable solutions

Development Refine system requirements Create solution description Build system

Verify and validate system

Production Produce systems Inspect and verify

Utilization Operate system to satisfy users’ needs

(19)

Roles of knowledge

(20)

(21)

Who are the

agents that

share the

knowledge?

• Humans – all system

stakeholders – engineers,

operators, workers …

• Robots – vehicles,

manipulators, …

• Machines – machine tools,

conveyors, …

• Computers – databases,

(22)

Knowledge vertebrates the life-cycle

The Team

The Project

(23)

Systems and Flows

(24)

MBSE

• This modern approach to system construction is

known as

Model-based Systems Engineering

• System models are the

bricks and mortar of

engineers

• Engineers model the

(25)

The engineering workflow

Real

world

Model

world

Pw

Pm

_Sm

Sw

Engineering

produces solutions

to problems

through a world of

(26)

Knowledge (models) vertebrates

the life-cycle

The Team

The MBSE Project (

Models

)

(27)

Knowledge (models) vertebrates

the life-cycle

The Team

The MBSE Project (

Models

)

(28)

Deep MBSE models

• Deep MBSE models can be considered “knowledge”

in the philosophical/Gettier sense: they are

justified

true belief

• Intelligent robots with

faithful models

–of

themselves, of others – can:

• Act better

(29)

(30)

Ontology goes beyond computer

systems and humans

• Ontology is a specification of a conceptualization.

• It can be formalised and

be actionable by robots

(exercised in the pursue of their objectives).

• In software, “specifications” are

not just “concepts”

but real things.

• In MBSE proceses

specifications are realisable

(i.e.

(31)

Challenges

(32)

Merging of knowledge domains

• Different kinds of tasks have been using different

forms of knowledge

• They shall be

integrated

to enable systemic

emergence

• This is –obviously – the main role of ontologies

• The meaning of “

shared conceptualization

” reaches

(33)

Subsystem

System

Subsystem

Supersystem

(34)

Standard(s) alignment

• Too many threads of

conceptualization out

there

(35)

CORA and others

• Narrow standards may be a risk

• It is nonsense to address a

subdomain in isolation.

• It is unpractical to

address a subdomain.

(36)

End product (robot) is not the

only thing to be delivered

5/10/18 Sanz / Sharing Knowledge in the Intelligent

Robot Life-cycle 36

End product is

not

the only issue

Operational

Products

System

End

Product

Development

Products

• • •

Production

Products

Test

Products

Deployment

Products

Training

Products

Support

Products

Disposal

Products

Subsystem

Enabling

Product Sets

Consists of

Specification Decomposition

End Product

Interface Specifications

System

Specification

End Product Specification Development Product Requirements • • • Production Product Requirements Test Product Requirements Deployment Product Requirements Training Product Requirements Support Product Requirements Disposal Product Requirements Subsystem Specification Enabling Product Interface Specifications Subsystem Specification System External Interface Specifications Subsystem Interface Specifications

Resources : TeamWork

System

Core Team

End

Product

Team

Development

Products

Team

• • •

Production

Products

Team

Test

Products

Team

Deployment

Products

Team

Training

Products

Team

Support

Products

Team

Disposal

Products

Team

Subsystem

Team

Subsystem

Team

System layering

System End

Product DevelopmentProducts

• • • Production Products Test Products Deployment Products Training Products Support Products Disposal Products Subsystem Subsystem System End

• • • Production Products Test Products Deployment Products Training Products Support Products Disposal Products Subsystem Subsystem • • • System End

• • • Production Products Test Products Deployment Products Training Products Support Products Disposal Products Subsystem Subsystem • • •

Layer N Building Block

Layer N+1 Building Blocks

System Layering

S y s te m

E n d P ro du ct D e ve lo pm e n tP ro d uc ts

P rod u c tio n P r o du ct s T e st P r o du ct s

D e pl oy m en t P ro d uct s

T r a in ing P rod uc ts

S up po r t P r od uc ts

Dis p o sa l P ro du cts S u bs yst e m S ub s yst e m

S y s t e m

E nd P r o du ct D e ve lo pm e n tP r o d uc ts

P rod u c tio n P r o du ct s T e st P ro du ct s

Tra in in g P r od u cts

S up p ort P rod uc ts

Dis p o sa l P ro du ct s

S y s te m

E n d P ro du ct

D e ve lo pm e n t P r o d uc ts

P r od uc tio n P ro du ct s T e st P ro du ct s

D e pl oy m en t P ro d u ct s

T ra in ing P rod uc ts

S up po rt P rod uc ts

Dis p o sa l P rod u cts S u bs yst e m S ub s yst e m

S y s te m

E n d P ro d uc t

Dev el op m e nt P rod u cts

P r o d uc t ion P rod uc ts T e st P rod uc ts

De plo ym e nt P rod uc ts T ra i nin g P ro d uc ts

S u pp o rt P ro d uct s

Di s po sa l P ro d uct s S u b sy st em S u bs y ste m

S y s te m

E n d P ro d uc t Deve l op m e ntP rod uc ts

P ro du c t ion P rod uc ts T e st P rod uc ts

De plo ym e nt P rod uc ts T ra i nin g P r o d uc ts

S u pp o rt P ro d uct s

Di s po sa l P ro d uct s

S y s te m

E n d P ro du ct

De ve lo pm e n t P ro d uc ts

P rod uc tio n P r o du ct s T e st P r o du ct s

D e pl oy m en t P ro du ct s T r a in ing P r od uc ts

S up po rt P r od uc ts

Dis p o sa l P rod u cts

S y s te m

E n d P rod uc t Deve l op m e ntP r od uc ts

P ro du c t ion P ro d uc ts T e st P rod uc ts

De plo ym e nt P r od uc ts T ra i nin g P ro d uc ts

S u pp o rt P ro d u ct s

Di s po sa l P ro d uct s S u bs y ste m S u bs y ste m

S y s te m

Dis p o sa l P ro du cts

S y s te m

E n d P rod uc t

Deve l op m e nt P r od uc ts

S y s te m

Dis p o sa l P ro du cts S u bs yst e m S ub s yst e m

S y s te m

E nd P ro du ct

D e ve lo pm e n t P r o d uc ts

P rod uc tio n P ro du ct s T e st P r o du ct s

Tra in in g P rod u cts

S up p ort P rod uc ts

Dis p o sa l P ro du ct s

S y s t e m

E n d P rod uc t Deve l op m e ntP r od uc ts

P ro du c t ion P rod uc ts T e st P rod uc ts

De p lo ym e nt P r od uc ts T ra i nin g P ro d uc ts S u pp o rt P ro d uct s

S y s te m

E n d P ro d uc t

Deve l op m e nt P rod uc ts

P ro du c t ion P ro d uc ts T e st P ro d uc ts

De p lo ym e nt P r od uc ts T ra i nin g P ro d uc ts

S u pp o rt P ro du ct s

Di s po sa l P r o d uct s

P rojec t B U s er or C ustom e r

D es ired S y stem

O ff-The-S helf/R euse E nd P roducts

P roject B’s T op-Layer B uilding B lock

B uilding B loc k D evelopm ents

B uild/C ode E nd P roducts P rojec t A

Stackholders at all levels

L a y e r 1 B u ild in g

B lo c k

L a y e r 2 B u ild i n g

B lo c k s

E n d P r o d u c t S p e c ifie d R e q u i r e m e n t s

L a y e r 3 B u il d in g

B lo c k s

L a y e r 4 B u ild in g

B lo c k s

L a y e r 5 B u il d in g

B lo c k s

D e s ig n F e e d b a c k A s s ig n e d

S p e c i fie d R e q u ir e m e n ts

O t h e r S ta k e h o ld e r s R e q u ire m e n ts

P r o je c t B (s e e F ig u r e 6 .2 . 1 a )

D e s ig n F e e d b a c k

O t h e r S ta k e h o ld e r s R e q u ir e m e n ts

O t h e r S ta k e h o ld e r s R e q u ir e m e n t s

O th e r S ta k e h o ld e r s R e q u ir e m e n ts

O th e r S t a k e h o l d e r s R e q u ir e m e n t s

D e s ig n F e e d b a c k

A c q u i re r D e s ir e d S y s t e m

D e v e lo p e r

E n d P r o d u c t S p e c ifie d R e q u ire m e n ts

E n d P r o d u c t S p e c ifie d R e q u ir e m e n t s

E n d P r o d u c t S p e c ifie d R e q u ir e m e n ts

E n d P ro d u c t S p e c i fie d R e q u ir e m e n ts A s s i g n e d

S p e c ifi e d R e q u ir e m e n t s A s s ig n e d

S p e c if ie d R e q u i r e m e n ts A s s ig n e d

S p e c ifi e d R e q u ir e m e n ts A s s ig n e d

S p e c if ie d R e q u i r e m e n t s

(37)

Further thougts

(38)

From heaven down to earth

• Models for MBSE of intelligent robots shall be

complete in many dimensions: domain, task,

abstraction level, etc.

• Shall be multilevel from an abstraction perspective: from

very abstract to fully concrete (actionable).

Multiresolutional MBSE is in need.

• Shall use heterogenous modeling technologies.

• Etc.

(39)

Not 3 but 7 Systems

(40)

1. Context System (S1) contains a Problem (P1)

2. Intervention System (S2) is intended to address P1

3. Realization System (S3) brings S2 into being

4. S2 is a constituent of S3

5. S3 needs to understand S1

6. S3 needs to understand the Modified Context System (S1’)

7. S3 may need to develop or modify the Sustainment System (S6)

8. Intervention System (S2) becomes Deployed System (S4)

9. S1 becomes the Modified Context System (S1’)

10. S4 is contained in S1’

11. S4 collaborates with one or more Collaborating Systems (S5)

12. S4 is sustained by Sustainment System (S6)

13. S4 may cause new Problem (P2)

Summary of

interactions

between these

seven systems:

(41)

A Broader Stance

• Autonomous Robot Ontologising (ontologists work)

is a domain-specific ontological activity that

requires a broader mental stance:

• Extended life-cycle

vision:

• E.g. Robots can exercise/actionate knowledge acquired

by others in the past or can produce knowledge to be

used by others in the future.

• Systemic

vision:

• Ontology work cannot be centered in the narrow

(42)

What are core concepts for this

work?

• System/Environment

• Structure/State

• Behavior

• Task

• Function

• Knowledge (Meta knowledge is critical)

(43)

“Function”

can vertebrate the three

systems along the whole life cycle

The Team

The Project

(44)

The concept of “function”

• The Function Ontology

- https://fno.io/spec/

• A

function

is an actionable thing that can have input

parameters, output, solve certain problems, and implement

certain algorithms.

• SEBOK

-https://www.sebokwiki.org/wiki/Function_(glossary)

• (1)

A system outcomes which contribute to goals or objectives.

To have a function, a system must be able to provide the

outcome through two or more different combinations of

elemental

behavior

.

(45)

Function in BFO

A function f is

(1) a realizable dependent continuant. Thus,

(2) it has a bearer, which is an independent

continuant, and

(3) it is of a type instances of which typically have

realizations; each realization is

a. a process in which the bearer is participant

(46)

Two (or more) views of function

• Structure/Behavior

• Goal/Effect

• How can different views of function be

captured

(47)

Other themes of relevance

• Continuous sytems

• BPMN and Knowledge-Intensive Processes (KIP)

• Knowledge objects - metaknowledge

• Cross-engineering MBSE

• Multiresolution modeling by functional lumpability

• Context (the rest of the MODEL)

(48)

References

• J. N. Martin. The seven samurai of systems

(49)