IE 5305 Cognitive Engineering: Advanced Cognitive Human Factors

05. Cognitive System and Cognitive Work

Sep 5, 2024

Changwon Son, PhD, CSP

Department of Industrial, Manufacturing, & Systems Engineering

Texas Tech University
Reading Material #1

2
Reading Material #1

Back in the days…

• Machines were not modern computers we use
now and involved processes were simple.

• Machines required more of human physical and

less of a human operator’s cognitive skills.

• MMS studies (now human-computer studies)

were focused on physical and mechanical
relationships.

3
Reading Material #1

4
Reading Material #1

5
“Intelligent Action”

6
Cognitive System

v Definition

§ A cognitive system is a collection of parts or elements that is able to

perform cognitive functions (e.g., perceiving, reasoning, deciding,
and problem-solving) to accomplish the system goals.

§ A complex socio-technical system is a cognitive system.

• Humans (‘socio-’) are one of the essential components of the system.

• Interactions between humans and other systems components enable a system operation.

7
Cognitive System

v Cognitive System Triad

World imposes demands
World/Work Domain
– goals to accomplish and
functions to carry out

Strategie
s
Field of

Af
Practice

for

ls
nd

oo
da

ma

s/T
nc
Machines/tools provide

De
es

ine
affordances to enable the
Ag

Agent develops strategies

ch
agent to implement the
en

Ma
to achieve the goals and strategies
t

execute the functions

8
Cognitive System

v World/Work Domain
World imposes demands
World/Work Domain
– goals to accomplish and
functions to carry out

Strategie
s
Field of

Af
Practice

for

ls
nd

oo
da

ma

s/T
nc
Machines/tools provide

De
es

ine
affordances to enable the
Ag

Agent develops strategies

ch
agent to implement the
en

Ma
to achieve the goals and strategies
t

execute the functions

9
Cognitive System Elements

v Cognitive System Triad

§ World or work domain

• Goal-means structure (e.g., goals-functions-physical actions)

• Hazards (i.e., potentials to cause harm)
• Temporal dynamics (i.e., time pressure)
• Uncertainty and risk (e.g., unpredictable states or consequence of an action)
• Complexity of interactions between system elements
• Coupling between system elements
• System constraints

10
World/Work Domain

v Goal-means structure
§ ‘Why’ – ‘What’ – ‘How’ relationship

More abstract Goals and Why does the system exist?

(high-level) objectives

Functions What functions need to occur to accomplish goals and objectives?

More concrete Physical How are the functions realized by physical actions?
(low-level) actions
11
World/Work Domain

v Hazards in a system
§ Hazard = Potential to cause harm to people, property, and environment

• Nuclear power generation – Radioactive materials (e.g., plutonium, uranium)

• Air transportation – High speed and high altitude of airborne flights

• Heavy manufacturing – Heavy materials and heavy equipment

• Hospital – Infection (virus, bacteria, germs), physical and mental stress

12
World/Work Domain

v Temporal dynamics
§ Time pressure is common in a complex socio-technical system

• ‘Faster, better, cheaper’ – Driving factors of modern industrial systems (e.g., Just-in-time)

§ Efficiency-Thoroughness Trade-Off (ETTO)

• Efficiency – an ability to produce a desired outcome without wasting materials, energy,
effort, and time

• Thoroughness – an ability to make an accurate decision or action with a large amount of

attention, care, and reasoning

13
World/Work Domain

v ETTO Principle
§ There is a trade-off between efficiency and thoroughness

Efficiency: “Time to do” Thoroughness: “Time to think”

Given time & resource

• Executing actions • Recognizing situation
• Implementing plans • Choosing options
• Faster actions • Careful actions
• Actions with a little effort • Actions after sufficient consideration
14
World/Work Domain

v ETTO Principle
§ When ‘efficiency’ is emphasized too much,
• Actions become poorly prepared or haphazardly executed.
• Actions are inappropriate for the situation.
• Intended outcomes are not produced.

§ When ‘thoroughness’ is emphasized too much,

• Actions are performed too late.

• There is little time to act (incomplete action or not taking actions at all).
• Intended outcomes are not produced.
15
World/Work Domain

v Uncertainty and Risk

§ Environments and system operating conditions are uncertain.

Environmental conditions in an airport Conditions of a patient admitted to an ER

16
World/Work Domain

v Complexity and Coupling

§ Complexity [Example – Air Traffic Control Center]

• Number of feedback loops and

interactions among system components

§ Coupling
• Degree to which “slack”/flexibility can be
tolerated in system interactions without
affecting system performance

• Tight coupling: little slack

• Loose coupling: much slack
17
World/Work Domain

v System constraints

§ Types of system constraints (which impact the design and operation of the system)
• Design constraints – limitations set by design (e.g., maximum speed of a passenger vehicle)

• Technical constraints – limitations set by the technology (e.g., maximum charging speed of
EV batteries)

• Economic constraints – limitations set by costs and budget (e.g., lowest airfare, lowest
medical fee)

• Legal constraints – limitations set by the existing laws and regulations (e.g., school teachers
cannot be armed against active shooters)

• Social constraints – limitations set by the society (e.g., not-in-my-backyard for nuclear plants)
18
In-class Discussion

v World/Work Domain
§ Example - Airport
Let’s discuss the following:
• Goal-means structure
• Hazards
• Temporal dynamics
• Uncertainty and risk
• Complexity
• Coupling
• System constraints
19
Cognitive System

v Agent
World imposes demands
§ Human agent World/Work Domain
– goals to accomplish and
§ Computerized agent functions to carry out
(e.g., AI)
§ Joint cognitive agent
Strategie
(human+computer)
s
Field of

Af
Practice

for

ls
nd

oo
da

ma

s/T
nc
Machines/tools provide

De
es

ine
affordances to enable the
Ag

Agent develops strategies

ch
agent to implement the
en

Ma
to achieve the goals and strategies
t

execute the functions

20
Cognitive System Elements

v Cognitive System Triad

§ Agent (Human and/or Computerized)

• Information processing
• Perception
• Memory and attention
To develop strategies to
• Skills and knowledge cope with problems in a
system
• Sources of error
• Sources of adaptations (or endogenous variability)
• Communication and coordination (among multiple agents)

21
Cognitive System Elements

v Skill-, Rule-, and Knowledge-based (SRK) Behavior

§ Skill-based behavior
• Highly-practiced, near-automated, and largely physical actions that require
little effort to monitor, reason, or establish decisions.
• Examples - Turning on/off light, typing with a keyboard, changing vehicle gear

§ Rule-based behavior
• Some degree of conscious reasoning and actions that require an application of
appropriate rules to given situations.
• Examples – Lowering AC temperature on an abnormally hot day, muting oneself in a
zoom meeting, pulling over one’s vehicle when an ambulance passes by
22
Cognitive System Elements

v Skill-, Rule-, and Knowledge-based (SRK) Behavior

§ Knowledge-based behavior
• High degree of conscious and effortful reasoning and actions to deal with
unfamiliar situations
• Examples – Water leaking from a ceiling, receiving a suspicious email (e.g.,
phishing email), vehicle crash in a place you are visiting for the first time

23
Cognitive System Elements

v Adaptive Process of a Cognitive System

(Re-)
§ A cyclic process of adaptation r m al Planning
no
“Ab avior”
• Human or computerized agents beh for
o r
are tasked with adapting to o n it ss
M ce
changes and anomalies Anomaly suc
Recogni- Corrective
• Adaptation is a key to sustaining a tion Di response
ag
system operation. s no
“U e arc stic
ne h
b e xpe
ha c
vio ted
r”
Diagnosis

24
Cognitive System

v Machines/Tools
World imposes demands
World/Work Domain
– goals to accomplish and
functions to carry out

Strategie
s
Field of

Af
Practice

for

ls
nd

oo
da

ma

s/T
nc
Machines/tools provide

De
es

ine
affordances to enable the
Ag

Agent develops strategies

ch
agent to implement the
en

Ma
to achieve the goals and strategies
t

execute the functions

25
Cognitive System

v Power of Representations by Machines and Tools

Example: Game of 15

Alternate picking numbers 1-9 until one player has a set of three that add to 15
(without picking the same number again)

1 2 3 4 5 6 7 8 9

Player A – 7, 3, 5 = 15

Player B – 8, 1, 6= 15
26
Cognitive System

v Power of Representations by Machines and Tools

1 2 3 4 5 6 7 8 9

27
Cognitive System

v Power of Representations by Machines and Tools

2 7 6

9 5 1

4 3 8

28
Cognitive System

v Two Gulfs between Agent and World

Machines

Gulf of
Evaluation
Machines

World/
Agent work domain
Gulf of
Execution

Strategies Goals and demands

29
Cognitive System

v Two Gulfs between Agent and World

§ Gulf of Evaluation (or Observability)
• A degree of difficulty of assessing the system state or demands of the world.

• Machines/tools act as a representation of the system or the work domain.

§ Gulf of Execution (Or Directibility)

• A gap between an agent’s action to accomplish the goals and the means to
execute such an action.

• Machines/tools act as an instrument or prothesis to help the agent execute the

action.
30
Cognitive System

v Machines/Tools

[Example – AC thermostat]
• Gulf of Evaluation?
- How can you know the
current status of air
conditioning of the house?

• Gulf of Execution?
- How can you adjust the
temperature of the house?

31
Cognitive System

v Machines/Tools
[Example – NPP Control Interface]
• Gulf of Evaluation?

• Gulf of Execution?

32
Cognitive System

v Human’s interaction with System

World/Work Domain § Human-system interaction becomes
increasingly ‘indirect’.

Strategie § Overdependence on machines and

s
technologies
Field of
Af

Practice
for

ls
§ Clumsy automation (autopilot)
nd

oo
da

ma

s/T
nc

De
es

ine
Ag

ch

§ Reduced user experience with the real

en

Ma
t

world (system mental model)

33
Any Questions?