Agent-based modelling for analysis of resilience in ATM Sybert - - PowerPoint PPT Presentation

▶
agent based modelling for analysis
SMART_READER_LITE
LIVE PREVIEW

Agent-based modelling for analysis of resilience in ATM Sybert - - PowerPoint PPT Presentation

Sep 26, 2022 282 likes •575 views

Agent-based modelling for analysis of resilience in ATM Sybert Stroeve, Tibor Bosse, Henk Blom, Alexei Sharpanskykh, Mariken Everdij SESAR Innovation Days 2013, Stockholm, Sweden Contents Resilience and the objective of MAREA Agent-based

slide-1
SLIDE 1

Agent-based modelling for analysis

  • f resilience in ATM

Sybert Stroeve, Tibor Bosse, Henk Blom, Alexei Sharpanskykh, Mariken Everdij SESAR Innovation Days 2013, Stockholm, Sweden

slide-2
SLIDE 2

Contents

Resilience and the objective of MAREA Agent-based modelling for analysis of resilience in ATM Development of a library of model constructs Integration and application agent-based model constructs Conclusions

2

slide-3
SLIDE 3

‘Resilience’

3

Google ngram diagram for the term ‘resilience’

Resilience Engineering for safety management Materials science Ecology Social sciences

slide-4
SLIDE 4

Resilience definitions

Socio-ecological systems (Folke, 2006)

 Capacity of a system to absorb

disturbance and re-organize while undergoing change so as to still retain essentially the same function, structure, identity and feedback

4

Resilience Engineering for ATM (Eurocontrol, 2009)

 The intrinsic ability of a system to

adjust its functioning prior to, during, or following changes and disturbances, so that it can sustain required operations under both expected and unexpected conditions

slide-5
SLIDE 5

Human role in resilience

Flexibility and system oversight by human operators in ATM are essential for efficient and safe operations in normal and rare conditions

5

Resilience Engineering emphasises the performance variability of human operators in normal and rare conditions

 Accounting for a broad range of

human factors

 Away from human error thinking

slide-6
SLIDE 6

6

Mathematical Approach towards Resilience Engineering in ATM (MAREA)

Aim To develop a mathematical modelling and analysis approach that allows to bring Resilience Engineering at work for the complex ATM system Focus on human performance

Humans dealing with uncertainties and non-nominal conditions

Psychological and organizational models

slide-7
SLIDE 7

Agent-based modelling of the ATM sociotechnical system for analysis of resilience

7

Model construct Human operator agent Model construct Model construct Human operator agent Model construct Model construct Technical system agent Model construct Model construct Technical system agent Model construct Model construct Model construct Environment

Agent-based model of a sociotechnical system Emergent properties

 Micro-level: behaviour

described by model constructs & interactions

 Macro-level: resilience

indicator for “required

  • perations are sustained”

 Macro-level properties

are the resultant of interacting micro-level properties

slide-8
SLIDE 8

Need for model constructs of disturbances and performance variability in the sociotechnical system

8

Model construct Human operator agent Model construct Model construct Human operator agent Model construct Model construct Technical system agent Model construct Model construct Technical system agent Model construct Model construct Model construct Environment

slide-9
SLIDE 9

9

Hazards as descriptions of disturbances and performance variability in ATM

Hazard = “Anything that may influence safety”

 Events / conditions / performance aspects  Humans / systems / environment  Interactions

NLR ATM Hazard Database

 ATM safety assessments  Hazard brainstorm sessions

– Pilots / Controllers / Experts – No analysis allowed

 4000+ hazards

slide-10
SLIDE 10

10

A set of generalized hazards

4000+ Selection of unique hazards 525 Generalization of hazards Development Validation Wrong waypoints in database Transponder sends wrong call-sign False alert of an airborne system Pilot mixes up ATC clearances Pilot validates without checking Alert causes attentional tunneling Controller has wrong SA about intent of aircraft Flight plans of ATC system and FMS differ Weather forecast is wrong Resolution of conflict leads to other conflicts Contingency procedures have not been tested 266 259

slide-11
SLIDE 11

Development of a library of model constructs

Identification of model constructs in three sources/phases

1.

NLR TOPAZ multi-agent dynamic risk modelling

2.

Agent system research at VU University Amsterdam

3.

Other sources Which model constructs can model the set of hazards?

11

slide-12
SLIDE 12

Library of model constructs

  • 1. TOPAZ MA-DRM
  • 2. VU agent system research
  • 3. Complementary

Human information processing Object-oriented attention Approach Multi-agent situation awareness Experience-based decision making Handling inconsistent information Task identification Operator functional state Group emotion Task scheduling Information presentation Confusion/ Surprise – Complex Procedures Task execution Safety culture Confusion/ Surprise – Changed Procedures Cognitive control mode Situation awareness with complex beliefs Deciding when to take action Task load Trust Access rights Human error Formal organisation Merging or splitting ATC sectors Decision making Learning Bad weather System mode Goal-oriented attention Weather forecast wrong Dynamic variability Extended mind Turbulence Stochastic variability Icing Contextual condition Influence of many agents on flight planning Uncontrolled aircraft

12

Initial model set (13) Final model set (38)

slide-13
SLIDE 13

Examples of model constructs: Multi-agent situation awareness

13

Multi-agent system:

             

,

identity state mode intent

j t k

SA of agent k at time t about agent j :

state

agent 1

SA

agent 2

Observation

SA

agent 1

SA

agent 2

Communication

SA

agent

decision

rules

Reasoning

SA updating processes:

slide-14
SLIDE 14

Examples of model constructs: System mode

14

Mode 1 Mode 2 Mode 3

  • Failure modes of technical systems
  • Normal working modes of technical systems
slide-15
SLIDE 15

Examples of model constructs: Operator functional state

15

External World Operator

Task Demands Situational Aspects Task Execution State Task Demands Environment State Actions Recovery Effort Experienced Pressure Generated Effort Provided Effort Effort Motivation Task Goals Processing Expertise Profile Personality Profile Basic Cognitive Abilities Critical Point Maximal Effort Exhaustion

slide-16
SLIDE 16

Examples of model constructs: Group emotion

16

Emotion

S SR R R

qS qR

slide-17
SLIDE 17

Examples of model constructs: Confusion / Surprise

17

slide-18
SLIDE 18

Matching model constructs with hazards: Mental simulation process

Mental simulation of all hazards by multiple analysts Result of a mental simulation per hazard:

 The relevant model constructs  Category: Well modelled / Partly modelled / Not modelled

Example hazard

 ‘Pilots do not react to controller call due to high workload’

18

  • 1. opportunistic
  • 2. call
  • 3. unimportant
  • 4. low

priority

  • 5. no response
slide-19
SLIDE 19

Matching model constructs with hazards (examples)

19

Hazards Models

Pilots do not react to controller call due to high workload

  • Task scheduling
  • Task execution
  • Cognitive control mode

Failure of GPS system

  • System mode

Pilot reports wrong position

  • Human error
  • Multi-agent SA

Pilot performance is affected due to alcohol, drugs or medication

  • Operator functional state
  • Safety culture
slide-20
SLIDE 20

Matching model constructs with hazards (examples)

20

Hazards Models

Confusion due to many sources that provide you with information

  • Multi-agent SA
  • SA complex beliefs
  • Trust
  • Confusion/Surprise (A)

Controller does not use alert system

  • Human error
  • EB decision making
  • Trust

Controller is frustrated with employer

  • Formal organisation
  • Group emotion
slide-21
SLIDE 21

Hazard modelling results

21

0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% 70.00% 80.00% 90.00% 100.00% Well modelled Partly modelled Not modelled

Initial

Model set:

Validation

Hazard set:

Final Final Development Development

slide-22
SLIDE 22

Top 15 of frequency of model constructs for hazard modelling (all hazards)

Rank Model construct Total No. Perc. 1 Multi-agent situation awareness 219 41.7% 2 System mode 118 22.5% 3 Human error 117 22.3% 4 Human information processing 95 18.1% 5 Task execution 57 10.9% 6 Dynamic variability 53 10.1% 7 Situation awareness with complex beliefs 50 9.5% 8 Operator functional state 49 9.3% 9 Stochastic variability 48 9.1% 10 Contextual condition 48 9.1% 11 Experience-based decision making 40 7.6% 12 Formal organisation 37 7.0% 13 Task scheduling 31 5.9% 14 Trust 31 5.9% 15 Confusion / Surprise (A) 28 5.3%

22

Initial models Additional models

slide-23
SLIDE 23

High-level integration of model constructs for human agents

23

Safety culture - awareness Situation awareness with complex beliefs Decision making Trust Deciding when to take action Confusion / Surprise

Environment of human agent

Human agent

Task scheduling Task identification

sensemaking deciding functional state task planning actuating

Information presentation Extended mind - perception Goal-oriented attention Object-oriented attention

sensing

Task execution Extended mind - effectuation Multi-agent situation awareness Cognitive control mode Group emotion - emotional state Task load Operator functional state Human error Learning Dynamic variability Stochastic variability Safety culture - interaction Group emotion - contagion

slide-24
SLIDE 24

Formalisation and simulation of integrated model constructs

 Illustrations of formalisation and simulation of integrated model

constructs are presented in MAREA deliverables and papers

 Formalisation and simulation of the model constructs are also

shown in various papers on TOPAZ multi-agent dynamic risk modelling and VU agent-based modelling applications

24

slide-25
SLIDE 25

25

ground control sector ground control sector runway control sector

Runway incursion scenario Multi-agent system Model in SDCPN syntax

Formalisation and integration of model constructs in TOPAZ multi-agent dynamic risk modelling (example)

slide-26
SLIDE 26

Formalisation and integration of model constructs in TOPAZ multi-agent dynamic risk modelling (example)

26

Risk results

How well can agents contribute to avoiding an accident in a runway incursion?

Monte Carlo simulation + speed-up

(Stroeve et al., 2013)

slide-27
SLIDE 27

Conclusions

An extended library of agent-based model constructs has been developed with a variety of human, environmental and

  • rganizational models

 From 13 model constructs to 38 model constructs  The constructs can well model about 92% of a large set of

hazards, and partly model about 7%

 Model construct “Multi-agent situation awareness” is used

mostly (42% of hazards) The model constructs have been integrated and it has been shows that they can be formalised Agent-based modelling supports analysis of resilience in ATM

 Practical feasibility has been shown by multi-agent dynamic risk

modelling for safety-focused resilience Future research: application of the enhanced model set

27

slide-28
SLIDE 28

28

Questions / Discussion

Further reading: http://complexworld.eu/wiki/MAREA Questions?