Theme 5 US workshop reportReport compiled by Loughborough University
IndexThe aim of this report is to provide an overview of the discussions and outputs of a workshop held at Stevens Institute of Technology, USA on 23rd – 24th May 2016.
The focus of the workshop was Theme 5 [Integration of socio/legal/governance models
within a modeling framework] within the TAMS4CPS project. 4 earlier workshops have
been held in US and Europe on other themes. Reports on these workshops can be found
at:
https://www.dropbox.com/sh/70503zelp15peqw/AAArt6CCpXqnuCQJTHoXFvPOa?dl=0.
This set of slides which forms the report contains the following information. Note that
some of the presentation slides are included in the slide pack for interest but the reader
is invited to consult whichever section is of relevance to them. The sections are self
standing and not dependant the one on the other: hyperlinks to individual sections are
provided on the next slide.
• Overview of the Theme 5 workshop: location and attendees
• Presentations on the EU vision for CPS and State of the Art wrt Theme 5
• Summary of workshop outputs: simulation and modelling capability gaps relevant to Theme 5
• Details of potential Dream Projects relevant to Theme 5
Index
1. Overview of Theme 5 workshop
2. Summary of EU vision for M&S for CPS
3. Theme 5: Modelling the Human Factor in CPS –issues and challenges
4. Key Areas identified for M & S for Theme 5
5. Workshop outputs: M & S capability gaps identified
6. Workshop outputs: Expansion of relevant Dream Projects
7. Workshop outputs: Key Discussion Points
8. Conclusions
9. Acknowledgements
1. Overview of Theme 5 US Workshop
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Overall Scope of TAMS4CPS Workshops
Overall aim of US Workshops:
To identify priority research and development needs for modelling and
simulation (M&S) for Cyber-Physical Systems (CPS) and provide key enablers
for trans-atlantic collaboration.
Workshop objectives:
• To share EU and US perspectives on the state of the art, strengths and opportunities, and
vision for M&S of CPS
• To identify and prioritize CPS M&S trends and drivers, markets, technologies and
enablers/barriers
• To elicit “test cases” for benchmarking CPS M&S products
• To determine research collaboration priorities and implementation plans for M&S of CPS
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Scope of Theme 5 Workshop
Integration of socio/legal/governance models
within modeling frameworks
Models of technical systems must necessarily make
assumptions about the environment in which the
resulting technical systems will operate, and these
systems will be at their most effective and efficient
when their links to society, its laws and people are
seamless.
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
5 Key Areas identified for Theme 5
• Modelling of human behavior and performance when individuals are interacting with CPS. This involves both lay people who have no interest in the system other than its performance as well as trained operators co-working within CPS, human augmentation, and novel interfaces such as exoskeletons
• Modelling of employment issues for people who are co-workers within CPS. This covers training, education, job design, inclusivity, etc.
• Modelling of decision and control within CPS. This addresses issues of the allocation of authority and responsibility; situation awareness, informed command and informed consent; etc.
• Modelling of external governance of CPS. This covers accountability, regulations both to assure compliance with legal aspects and to create a ‘level playing field’ for CPS within society.
• Modelling of societal aspects within business models. This enables the exploration of Corporate Social Responsibilities, Responsible Research and Innovation, and other aspects such as integrity, trust and acceptability.
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Agenda: Day 1Session 1 Mutual introductions: EU and US perspectives on the state of the art, EU and US
strengths & opportunities, EU and US vision
09:00-
09:45
Welcome, introductions and overview of TAMS4CPS LU
09:45-
10:15
EU vision and capability in M&S for CPS LU
10:15-
11:00
EU State of the Art in Theme 5 M&S for CPS LU
11:00-
11:30
Coffee
11:30-
12:30
Round Table discussion, participants mini-presentations on US vision,
state of the art and strengths & opportunities
All
12:30-
13:30
Lunch
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Agenda: Day 1Session 2 Exploration and identification of trends & drivers; needs / markets; opportunities;
technologies (either current or needed); research directions; enablers / barriers
13:30-
13:45
Process for the afternoon, reminder of areas for consideration LU
13:45-
14:30
Gap analysis, Requirements, Impact - M&S Human Behaviour: potential
topics and priorities + 15 mins brief presentation to participants
Groups
14:30-
15:15
Gap analysis, Requirements, Impact - M&S Employment Issues: potential
topics and priorities + 15 mins brief presentation to participants
Groups
15:15-
15:30
Coffee
15:30-
16:15
Gap analysis, Requirements, Impact - M&S Decision & Control: potential
topics and priorities + 15 mins brief presentation to participants
Groups
16:15-
16:45
Gap analysis, Requirements, Impact – M&S External Governance of CPS:
potential topics and priorities
Groups
16:45-
17:15
Gap analysis, Requirements, Impact - M&S Societal Aspects within business
models: potential topics and priorities
Groups
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Agenda Day 2Session Timing Overview
1 May 16
a.m.
Mutual introductions: EU and US perspectives on the state of the art,
EU and US strengths & opportunities, EU and US vision
2 May 16
p.m.
Exploration and identification of trends & drivers; needs / markets;
opportunities; technologies (either current or needed); research
directions; enablers / barriers
3 May 17
a.m.
Identification of research collaboration priorities and planning the
way forward. Identification of Dream Projects.
4 May 17
p.m.
Elicitation and development of test cases: Test case specifications will
be developed
5 May 17
p.m.
Wrap –up and forward planning
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Location and invited Participants
• Stevens Institute of Technology
• Professor Dinesh Verma
• Professor Greg Vesonder
• Dr Babrak Heydari
• Loughborough University
• Professor Carys Siemieniuch
• Dr Murray Sinclair
• Dr Luminita Ciocoiu
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Scene setting
It was felt to be beneficial to establish some common terms and understanding. So the following areas were briefly discussed. The same terms were used for the previous 4 workshops.
• Definitions of Cyber-Physical Systems [CPS]
• Classifications of CPS
• Characterisations of modelling & simulation
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Cyber-Physical Systems:
EU refer to ICT systems (sensing, actuating, computing, communication, etc.) embedded in physical objects, interconnected (including through the Internet) and providing citizens and businesses with a wide range of innovative applications and services (EC, 2013)
US can be described as smart systems that encompass computational (i.e., hardware and software) and physical components, seamlessly integrated and closely interacting to sense the changing state of the real world. These systems involve a high degree of complexity at numerous spatial and temporal scales
and highly networked communications integrating computational and physical components (Energetics Inc., 2013)
What is a Cyber-Physical System?
EC, 2013. European Commission. Cyber-Physical Systems: Uplifting Europe’s Innovation Capacity. Report from the Workshop on Cyber-Physical Systems: Uplifting
Europe’s Innovation Capacity, 29th – 30th October 2013, Brussels. December, 2013.
Energetics Inc., 2013. Foundations for Innovation in Cyber-Physical Systems, US Dept. Commerce, Washington DC, US: National Institute of Standards and Technology.
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
CPS can be characterised as follows (CyPhERS D5.2):
• Physical vs Embedded vs IT dominated
• Single Domain vs Cross-Domain
• Open vs Closed
• Level of autonomy
• Level of adaptability
• Distributed vs Centralised control
• Allocation of governance
• Single jurisdiction vs cross-jurisdiction
• Human In/Outside the Loop
• Degree of Integration
Classifying CPS
CyPhERS D5.2. Martin Törngren, Saddek Bensalem, María Victoria Cengarle, John McDermid, Roberto Passerone and Alberto Sangiovanni-Vincentelli. CPS: Significance,
Challenges and Opportunities. Document Number: D5.2. Technical report: http://www.cyphers.eu/sites/default/files/D5.2.pdf, 2014.
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Modelling:
• the activity of creating models (Fitzgerald et al., 2014)
A model:
• a partial description of a system, where the description is
limited to those components and properties of the system
that are pertinent to the current goal (COMPASS D11.3)
Modelling
John Fitzgerald, Peter Gorm Larsen, and Marcel Verhoef (Eds.). Collaborative Design for Embedded Systems: Co-modelling and Co-simulation. Springer-Verlag Berlin
Heidelberg, 2014.
COMPASS D11.3. Convergence Report 3; Document Number: D11.3. Technical report, http://www.compass-research.eu, October 2014.
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
• Level of modelling rigour: formal, semi-formal, informal
• Level of abstraction:
• physical, abstract (descriptive, analytical or hybrid)
• software-specific, platform-specific, platform-independent
• System aspects of interest: ICT, mechanical, human, etc.
• System lifecycle phase: design, implementation, operation, maintenance, disposal, etc.
• Modelling language characteristics:
• deterministic, non-deterministic
• continuous, discrete
• probabilistic, non-probabilistic
Classifying models
Extended and adapted from: Systems Engineering Body of Knowledge (SEBoK), http://sebokwiki.org/wiki/Types_of_Models (accessed April 2015).
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
A general definition:
• the imitation of the operation of a real-world process
or system over time (CPS-VO)
In the context of modelling:
• a model that behaves like a given system when
provided a set of controlled inputs (ISO/IEC/IEEE
24765:2010)
• symbolic execution of a model (Fitzgerald et al., 2014)
Simulation
CPS-VO. Cyber-Physical Systems Virtual Organisation. Tagcloud. Available online: http://cps-vo.org/tagadelic, accessed April 2015.
International Organisation for Standardisation. ISO/IEC/IEEE 24765:2010: "Systems and software engineering - Vocabulary", 2010.
J. Fitzgerald, P. G, Larsen, and M. Verhoef (Eds.). Collaborative Design for Embedded Systems: Co-modelling and Co-simulation. Springer-Verlag Berlin Heidelberg, 2014.
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Simulations can be classified according to the models
being executed, but also simulations can be:
• steady-state or dynamic
• local or distributed
• simulated or co-simulated
• live, virtual, or constructive
• software-in-the loop, hardware-in-the-loop,
network-in-the-loop
Classifying simulation
Extended and adapted from: Systems Engineering Body of Knowledge (SEBoK), http://sebokwiki.org/wiki/Types_of_Models (accessed April 2015).
2. Summary of EU vision for M&S for CPSDr Murray Sinclair
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
An overview of the current EU vision for CPS was presented to the US
participants in order to set the overall context for the workshop form the
EU perspective. This covered the following elements:
• Extracts from the CPS week held in Vienna in April 2016
• Key Technologies relevant to CPS driving the overall shift from product
to services
• Summary of EU initiatives for the Digitisation of Industry and the
challenges associated with this
• Overview of the Artemis Industry Association and its 2016 Strategic
Research Agenda relating to CPS
• Key elements of Horizon 2020 research programme where the focus is
CPS and associated research challenges
EU Vision for CPS
3. Theme 5: Modelling the Human Factor in CPS – issues and challengesDr Murray Sinclair and Professor
Siemieniuch
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Basic message
The inevitable human factor in CPS� Humans will be essential within the CPS ecosystem
for the foreseeable future
� Because CPS will reach deep into society, we must engineer socio-technical CPS, not just technical CPS
� Autonomy and entanglement with society entails that CPS are engineered for legal compliance, and to demonstrate governance and ethical behaviour
� Engineers will deliver these systems; nobody else is competent to do so.
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Contents
� A few fundamentals about humans and their
capabilities
� Some comments on legal issues
� Roles of humans within CPS
� Comments about human-CPS interfaces
� The bigger picture: sustainability
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
The ‘NIF’
Data/Object Model Interoperability
Connectivity & Network Interop.
Physical Interoperability
Semantic/Information Interoperability
Knowledge/Awareness of Actions
Aligned Procedures
Aligned Operations
Harmonized Strategy/Doctrines
Political or Business ObjectivesLa
yers
of
Inte
rop
era
bil
ity
Network
Transport
Information
Services
People &
Processes
& Applications
© NCIOC, https://www.ncoic.org/technology/deliverables/nif/
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
People and machines are different,
and also complementary
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
People within a CPS
� For functionality; for tasks that are not well-defined, or
happen in changing environments, or surprises are likely
� For system resilience, agility, and adaptability in a world of complexity and change; c.f. Heracleitus, BCE
500: “you cannot step in the same river twice”
� For governance, responsibility, & accountability; legal aspects, where CPS meet society
� For strategy, social responsibility, and ethical behaviour; appropriate, efficient, effective & robust CPS
behaviour within society over their lifecycles
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Role of people
� Informed command: UK MoD: “the last person to give
a command is responsible for the consequences.”
� Informed consent: “Yes/No/Wait”, e.g. in response to
automated suggestions
� Situation awareness: “the perception of environmental
elements with respect to time and/or space, the comprehension of their meaning, and the prediction of their status in the future.”
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
(Re-)defined systems
“A community of people,
pursuing a common purpose,
supported by a network of peripherals.”
S.V. Deal, 2008
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Human characteristics� Evolved over 800 million generations from the first
amoeba to what we are today
� Very good at sensing, perceiving and making use of the environment – dangers, patterns, affordances,
‘scaffolding’ (“The way is long if one follows precepts, but short and helpful if one follows patterns” (Seneca AD65))
� Social animals; rule-following, responsible, emotional
intelligence, ethics, team-working and trust
� Variable; no 2 humans alike – both a strength and a
weakness
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Human variability“In fact, humans differ enormously from machines, in that they areinherently variable and unreliable in their detailed behaviour,while simultaneously being hyper-adaptable and metastable intheir overall behaviour because they perceive and correct theirown errors.” (T.B. Sheridan, 2002)
� Well-designed jobs to achieve objectives
(meaningful tasks, safe operations, satisfying jobs)
� Education & Training to know processes, to understand
constraints and to minimise bad decisions
� Sufficient time to decide and execute actions correctly,
and to realise and retrieve wrong actions
The variability that remains provides governance, responsiveness, robustness, resilience, and agility (and is a problem for M&S)
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
On accountability and legal issues
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Legal Issues
� Main problem is to determine ‘good’ CPS behaviourin relation to ‘hard’ law and ‘soft’ law (codes of practice, guidelines, standards, etc.)
� ‘Good’ behaviour is situation-dependent
� Example of ‘Assistive living’ (including a robot) to highlight some of the legal aspects, and to illustrate the M&S problems
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
‘Assisted living’ (UK)
RefugeeRobot
Accommodation
Care network
Health delivery system
Local practice
Social Services
Robot OEM
Maintenance network
Robot network
services
medical
issues
Lease, supportSupply chain
Supply chain
Supply chain
maintenance
Care &
concern
Design
issues,
etc.
medical issues
Management
issues
Management
issues
Management
issues
Management
issues
Rest of family
Care &
concern
Management
issues
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Legal issues
� Develop robot capabilities in perception, individuation and prediction
� Develop legal frameworks and business models for open Systems-of-Systems that include robots (& other bots)
� Develop (run-time) verification, validation & certification processes for open SoS
� Develop robot ethics via simulation of scenarios that reflect business models
Necessary steps to ‘uncage’ robots for societal use:
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
‘Personhood’To be a responsible person in society:
� Sentience – conscious experience, pleasure & pain
� Emotionality – happy, sad, angry, loving, etc.
� Reasoning – solve new problems
� Communication – messages of many types and contents
� Self-awareness – self & place in social group
� Moral agency – regulate actions by moral & ethical principles
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Accountability (Asaro)
� For a robot to be held responsible for its acts, it must have ‘moral agency’
� In law, moral agency implies personhood, except emotionality – sociopaths are held responsible for their actions
� Robots are not likely to attain full moral agency; like pets, they will fall under ‘agency’ laws. The robot stakeholders will be responsible
� If robots attain a level of moral agency, we may need the ‘slave laws’ of Rome, or the Confederacy
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
People as stakeholders in CPS ecosystems
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
CPS in manufacturing
Virtual Object (milling
machine)
VO(robot arm)
VO(sensor)
VO(turningcentre)
VO(wifi)
VO(roadmap)
VO(forklift)
VO(truck)
Virtual Object(group attributes)
Business applications suite
Virtual Object(group attributes)
Market segment(consumers)
Market segment(businesses)
Composite Virtual Object(market opportunity)
Composite Virtual Object(transportation)
Composite Virtual Object(manufacturing cell)
Entrepreneur
INTERNETOF
THINGS
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Imperfections…• ”… it is worth noting that the findings from Capers Jones
and others indicate that deployed software systems contain
approximately 0.4 latent faults per function point. To our
knowledge, this indicator of dismal [operational availability]
has improved only about three-fold in the last four
decades." (Ring and Madni 2015)
• “In one case, we observed an outsourced application with
120 COTS products, 46% of which were delivered in a
vendor-unsupported state.” (Yang, Boehm et al. 2005)
• “Due to the large scale and the complexity of systems of
systems, the occurrence of failures is the norm in CPSoS”
(CPSoS report D2.4 (2015)).
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
IoT
Virtual Object (milling
machine)
VO(robot arm)
VO(sensor)
VO(turningcentre)
VO(wifi)
VO(roadmap)
VO(forklift)
VO(truck)
Virtual Object(group attributes)
Business application suite
Virtual Object(group attributes)
Market segment(consumers)
Market segment(businesses)
Composite Virtual Object(market opportunity)
Composite Virtual Object(transportation)
Composite Virtual Object(manufacturing cell)
Entrepreneur
INTERNETOF
THINGS
A ‘co-worker’, working with robots, smart systems, etc.
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
The need for MBSE“CPS generally support … critical processes, making it impossible to turn off the system to make changes … requiring (re-)configuration, (re-)deployment, (de-)commissioning, update, or enhancement during runtime.” (B. Schätz (2014)).
redesign,reconfiguration
models
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
The CPS Co-Worker
co-worker
redesign,reconfiguration
models
INTEGRATEDSYSTEMS
• Co-worker• Emergence
discoverer• Resilience
manager• Designer
experience & understanding
appropriateskills
roles & job design
strategiesWORLD &CULTURE
market
education
Kn. deployment
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Overview of current ‘Human’ models
� DODAF/ MODAF/NAF – while they may claim ‘human views’, these are
under-specified
� UK: DEF STAN 00-250 (20) – HF for designers of systems; covers many
aspects, but defence-oriented & does not integrate aspects well; omits
cultural/social aspects
� US: MANPRINT – Ditto (?replaced?)
� Other models:
• IPME (MicroSaint model of operations, includes an Operator model;
Dahn/Laughery 1997)
• ORSIM (system dynamics model of operation of Nuclear PP; Chu,
2006 )
• IMPRINT/ACT-R (cognitive perspective of a task network; Lebiere et
at, 2002)
• MIDAS-5 (NASA model of NextGen ATC; Gore et al, 2011)
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Modelling human-CPS interfaces
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Generic CPS interface
Cyber-Physical SystemInterface entities• Other co-workers
• Mobile devices
• Stationary devices
• Smart applications
• CPS context
• CPS environment
• CPS prognostics
co-worker
Situation
Awareness
Presentation
interface
Action
interface
INTERFACEVisualAuditoryHaptic
Transfers of MEI:
Mass, Energy &
Information
(Big Data issues)
VerbalHapticGestureLocationPostureMovement
Informed Consent,Informed Command
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Interface questions
Awareness questions
� What's happening?
� What isn't happening?
� Why are you telling me this?
� What does this mean?
� Is this answer trustworthy?
� What if ...... ?
� Remind me, how do I find ……?
� Where is ...... ?
Action questions
� What did I do?
� How do I undo this?
� How do I change this?
� How do I reach ....?
� What do I do next?
� What can I do next?
� How do I do it?
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Inclusive job design:
• Responsibilities
• Authority to act
• Job scope
• Mgmt support
• Goal setting & alignment
• Team working
• Training
• Trust
DEFAULT STATE
Supervise:
• maintain
situation
awareness,
readiness
level
OBSERVE
Detect event:
• Sampling rate
• Command or
alarm or
process event?
ORIENT
Characterise
event:
• Recognise
pattern(s)
DECIDE
Choose
response:
• NDM model
OR
• follow
procedure
ACT
Execute
response:
• Communicate
• Act
• Observe
effect(s)
(on completion of act,
return to default state)
Knowledge
state
• prior
experience
• Training
• process
experience
(capture
experience)
Execution process
Worker satisfaction
HRM aspects:
• Org. culture
• Mgmt style
• Pay
• Terms & conditions
• Governance
• Promotion
• Employability
Task design:
• Wkspace layout
• inclusivity
• Sit. awareness
• Robot roles
• Interfaces, IT
• Quality, Safety
• Procedures
• AR, manuals
• Tools, controls
• Assistance
Measured worker
satisfaction:
• Motivation
• Commitment
• Enthusiasm
• Continuity
• Willingness to learn
• Identity with company
CPS-Coworker interface
• Design of process interfaces
• Process support for situation
awareness for operator(s)
• Simulation facilities
• Use of AR, vision,
voice, sound &haptics
• Validation of commands
against capabilities, status
• Trust envelope for each
operator
• Propensity for error
• Performance
• Resilience
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Global sustainability issues
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Global issues
� Population demographics, especially growth and aging
� Food security
� Energy security
� Resource utilisation and re-utilisation
� Emissions and global climate
� Community security and safety
� Transportation
� Globalisation of economic and social activity
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Interactions of issues
Unwanted
outcomes
Population
demographics
Materials
depletion
Community
security &
safety
Transportation
systems
Economic
globalisation
Energy
security
Food
security
Emissions
‘wicked problems’;
‘normal’ accidents
climate
change
climate
change,
toxins
shortages
Increased
output
Increased
demand
Increased
demand
Increased demand,
extended networks
Increased
demand
Increased
demand
Increased
demand
Increased demand,
increased waste
Increased
demand
Increased
demand
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Addressing issues
Unwanted
outcomes
Population
demographics
Materials
depletion
Community
security &
safety
Transportation
systems
Economic
globalisation
Energy
security
Food
security
Emissions
‘wicked problems’;
‘normal’ accidents
climate
change
climate
change,
toxins
shortages
Increased
output
Increased
demand
Increased
demand
Increased demand,
extended networks
Increased
demand
Increased
demand
Increased
demand
Increased demand,
increased waste
Increased
demand Crop
science,
conservatio
n of waste
Systems
engineering,
resilience
renewables
CCS
CCS
Systems
engineering,
resilience
Changes to social
perceptions
‘circular’
manufacturing,
recycling
Increased
demand
Changes to social
perceptions
Non-carbon
technology
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Conclusions
� Humans will be essential within CPS ecosystems for the foreseeable future
� Because CPS will reach deep into society, we must engineer socio-technical CPS, not just technical CPS
� Modelling and simulation at many levels will be required, and we need tools for this
� Engineers will deliver these CPS; nobody else is competent to do so. This requires added skills and visions for engineers
� These visions need explaining for the public and for politicians to comprehend what is coming
4. Key Areas identified for M & S for Theme 5Dr Murray Sinclair and Professor
Siemieniuch
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Overview of Theme 5
Integration of socio/legal/governance models within modeling frameworks
Models of technical systems must necessarily make assumptions about the environment in which the resulting technical systems will operate, and these systems will be at their most effective and efficient when their links to society, its laws and people are seamless.
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
1. Modelling of human behavior and performance when individuals are interacting with CPS. This involves both lay people who have no interest in the system other than its performance as well as trained operators co-working within CPS, human augmentation, and novel interfaces such as exoskeletons
2. Modelling of employment issues for people who are co-workers within CPS. This covers training, education, job design, inclusivity, etc.
3. Modelling of decision and control within CPS. This addresses issues of the allocation of authority and responsibility; situation awareness, informed command and informed consent; etc.
4. Modelling of external governance of CPS. This covers accountability, regulations both to assure compliance with legal aspects and to create a ‘level playing field’ for CPS within society.
5. Modeling of societal aspects within business models. This enables the exploration of Corporate Social Responsibilities, Responsible Research and Innovation, and other aspects such as integrity, trust and acceptability.
5 key sub-areas identified
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
1. Modelling of human behavior and performance when individuals are interacting with
CPS. This involves both lay people who have no interest in the system other than its
performance as well as trained operators co-working within CPS, human augmentation,
and novel interfaces such as exoskeletons
co-worker
redesign,reconfiguration
models
INTEGRATEDSYSTEMS
• Co-worker• Emergence
discoverer• Resilience
manager• Designer
experience & understanding
appropriateskills
roles & job design
strategiesWORLD &CULTURE
market
education
Kn. deployment
Area 1: Human Behaviour and Performance
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Cyber-Physical SystemInterface entities• Other co-workers
• Mobile devices
• Stationary devices
• Smart applications
• CPS context
• CPS environment
• CPS prognostics
co-worker
Situation
Awareness
Presentation
interface
Action
interface
INTERFACEVisualAuditoryHaptic
Transfers of MEI:
Mass, Energy &
Information
(Big Data issues)
VerbalHapticGestureLocationPostureMovement
Informed Consent,Informed Command
Area 1: Human Behaviour &Performance- the Human /CPS interface
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Inclusive job design:
• Responsibilities
• Authority to act
• Job scope
• Mgmt support
• Goal setting & alignment
• Team working
• Training
• Trust
DEFAULT STATE
Supervise
:
• maintain
situation
awareness
, readiness
level
OBSERVE
Detect event:
• Sampling
rate
• Command or
alarm or
process
event?
ORIENT
Characterise
event:
• Recognise
pattern(s)
DECIDE
Choose
response:
• NDM model
OR
• follow
procedure
ACT
Execute
response:
• Communicate
• Act
• Observe
effect(s)
(on completion of act,
return to default state)
Knowledge
state
• prior
experience
• Training
• process
experience
(capture
experience)
Execution process
Worker satisfaction
HRM aspects:
• Org. culture
• Mgmt style
• Pay
• Terms & conditions
• Governance
• Promotion
• Employability
Task design:
• Wkspace layout
• inclusivity
• Sit. awareness
• Robot roles
• Interfaces, IT
• Quality, Safety
• Procedures
• AR, manuals
• Tools, controls
• Assistance
Measured worker
satisfaction:
• Motivation
• Commitment
• Enthusiasm
• Continuity
• Willingness to learn
• Identity with company
CPS-Coworker interface
• Design of process interfaces
• Process support for situation
awareness for operator(s)
• Simulation facilities
• Use of AR, vision,
voice, sound &haptics
• Validation of commands
against capabilities, status
• Trust envelope for each
operator
• Propensity for error
• Performance
• Resilience
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Area 2: Employment -humans as co-workers
2. Modelling of employment issues for people who are co-workers within CPS. This covers training, education, job design, inclusivity, etc.
co-worker
redesign,reconfiguration
models
INTEGRATEDSYSTEMS
• Co-worker• Emergence
discoverer• Resilience
manager• Designer
experience & understanding
appropriateskills
roles & job design
strategiesWORLD &CULTURE
market
education
Kn. deployment
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Inclusive job design:
• Responsibilities
• Authority to act
• Job scope
• Mgmt support
• Goal setting & alignment
• Team working
• Training
• Trust
DEFAULT STATE
Supervise
:
• maintain
situation
awareness
, readiness
level
OBSERVE
Detect event:
• Sampling
rate
• Command or
alarm or
process
event?
ORIENT
Characterise
event:
• Recognise
pattern(s)
DECIDE
Choose
response:
• NDM model
OR
• follow
procedure
ACT
Execute
response:
• Communicate
• Act
• Observe
effect(s)
(on completion of act,
return to default state)
Knowledge
state
• prior
experience
• Training
• process
experience
(capture
experience)
Execution process
Worker satisfaction
HRM aspects:
• Org. culture
• Mgmt style
• Pay
• Terms & conditions
• Governance
• Promotion
• Employability
Task design:
• Wkspace layout
• inclusivity
• Sit. awareness
• Robot roles
• Interfaces, IT
• Quality, Safety
• Procedures
• AR, manuals
• Tools, controls
• Assistance
Measured worker
satisfaction:
• Motivation
• Commitment
• Enthusiasm
• Continuity
• Willingness to learn
• Identity with company
CPS-Coworker interface
• Design of process interfaces
• Process support for situation
awareness for operator(s)
• Simulation facilities
• Use of AR, vision,
voice, sound &haptics
• Validation of commands
against capabilities, status
• Trust envelope for each
operator
• Propensity for error
• Performance
• Resilience
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Area 3: Decision Making and Control (1)
3. Modelling of decision and control within CPS. This addresses issues of the allocation of authority and responsibility; situation awareness, informed command and informed consent; etc.
� Includes co-working with autonomous devices, covered by
‘Agency law’.
� Autonomous devices necessarily must learn;
environment/navigation, performance polishing, strategy
improvement. Each of these produce behaviour change, of
increasing scope. Thus, they become ‘distanced’ from their
designer, and also their co-worker(s)
� For critical functions, how to perform V&V on devices that
learn? M&S seems the only viable answer.
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
� Do we allow autonomous devices to transmit learning
between them? If so, M&S again?
� Do we implement rules to restrict autonomy? e.g.
Sheridan, level 5: “Device offers a suggestion and executes
if human approves”; level 6: “Device allows human a
restricted time to veto, else executes”
� How do we model the dynamic flow and exercise of
authority and accountability within a CPS network? And
the constraints on this flow and exercise caused by
contracts?
Area 3:Decision Making and Control (2)
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Inclusive job design:
• Responsibilities
• Authority to act
• Job scope
• Mgmt support
• Goal setting & alignment
• Team working
• Training
• Trust
DEFAULT STATE
Supervise
:
• maintain
situation
awareness
, readiness
level
OBSERVE
Detect event:
• Sampling
rate
• Command or
alarm or
process
event?
ORIENT
Characterise
event:
• Recognise
pattern(s)
DECIDE
Choose
response:
• NDM model
OR
• follow
procedure
ACT
Execute
response:
• Communicate
• Act
• Observe
effect(s)
(on completion of act,
return to default state)
Knowledge
state
• prior
experience
• Training
• process
experience
(capture
experience)
Execution process
Worker satisfaction
HRM aspects:
• Org. culture
• Mgmt style
• Pay
• Terms & conditions
• Governance
• Promotion
• Employability
Task design:
• Wkspace layout
• inclusivity
• Sit. awareness
• Robot roles
• Interfaces, IT
• Quality, Safety
• Procedures
• AR, manuals
• Tools, controls
• Assistance
Measured worker
satisfaction:
• Motivation
• Commitment
• Enthusiasm
• Continuity
• Willingness to learn
• Identity with company
CPS-Coworker interface
• Design of process interfaces
• Process support for situation
awareness for operator(s)
• Simulation facilities
• Use of AR, vision,
voice, sound &haptics
• Validation of commands
against capabilities, status
• Trust envelope for each
operator
• Propensity for error
• Performance
• Resilience
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Area 4: External governance of CPS
4. Modelling of external governance of CPS. This covers
accountability, regulations both to assure compliance with legal
aspects and to create a ‘level playing field’ for CPS within society.
� Cybersecurity: protection, privacy, safety & security of data, IP,
strategy, etc.
� Physical safety & security of the CPS’ people, locations, processes and
operations
� Traceability of authority and control in operations
� Proof of real-time adequacy of the sensing-to-action decision sequence
� M&S for evaluating in real time the alternative action paths to deliver
the combination of ‘good ethical practice’ and ‘good management
practice’
� Gamification of ‘trust’ within business operations (Доверяй, но
проверяй: ‘trust, but verify’)
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Area 5: Societal Aspects
5. The modeling of societal aspects within business models. This
enables the exploration of Corporate Social Responsibilities,
Responsible Research and Innovation, and other aspects such as
integrity, trust and acceptability.
� M&S for cultural drivers of behaviour: individual, group,
professional, organisational, political, national
� M&S for the sustainability agenda: energy, water, waste,
materials, Big Data for control, etc.
� M&S as an aid to the ‘Responsible Research & Innovation’
(RRI) agenda; e.g. nano-materials, MEMs in health care,
balancing open-ness to future stake-holders while
maintaining security of data and intentions
5. Workshop outputs: M & S capability gaps identified
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
M & S Capability gaps 1 & 2
M & S Reqt
no
Sub M & S
Reqt NoDescription of Gap in M & S Capability Link to Dream Project Comments
1 Model of CPS Legislative Framework
✔Characterise and Improves Entry
and Use of CPS
✔Autonomous collection of
environmental data via drones
✔Assisted Living
1.1
M & S of legal/ethical aspects of use of
non-human autonomous intelligent
agents
✔Characterise and Improves Entry
and Use of CPS
✔Real-time system modelling
including human behaviour
✔Autonomous collection of
environmental data via drones
✔Assisted Living
2
A ‘gamification’ approach to modelling
to include responses of/consequences
for society and for co-workers within
CPS (i.e. CSR, RRI aspects)
✔Characterise and model dynamic
human interaction with CPS
✔Integration and interoperability
models and approaches
✔Real-time system modelling
including human behaviour
✔Assisted Living
Conserving house energy example
Change of human behaviour
Use of behavioural techniques to reinforce
desired behaviour
Taps into the competitive nature
2.1
Predicitive M & S of impact of
increased automation on employment,
processes, roles, organistaional
structure etc
✔Characterise and model dynamic
human interaction with CPS
✔Educating CPS designers, operators,
maintainers, co-workers
No internal feed-back to update model
parameters as model is operated
2.2
Pedicitive M & S of impact of increased
autonomy on employment, processes,
roles, organistaional structure etc
✔Characterise and model dynamic
human interaction with CPS
✔Educating CPS designers, operators,
maintainers, co-workers
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
M & S Capability gaps 6/7/8/9
M & S Reqt
no
Sub M & S
Reqt NoDescription of Gap in M & S Capability Link to Dream Project Comments
6M&S of impact of different
organisational structures
✔Characterise and model dynamic
human interaction with CPS
✔Real-time system modelling
including human behaviour
✔Assisted Living
Need to be able to create models at different
levels. The majority of models are at a
semantic level and are static rather than
dynamic
7
Models of different data/
communication links between humans
and non-human intelligent agents
✔Characterise and model dynamic
human interaction with CPS
✔Real-time system modelling
including human behaviour
✔Assisted Living
Current data transfer links {pipes} are small.
Response times are differents for various CPS
subsystems
8
M&S of human - machine, machine to
machine interactions, individually and
collectively
✔Characterise and model dynamic
human interaction with CPS
✔Assisted Living
No internal feed-back to update model
parameters as model is operated
9
Meta-level tools to (1) decide required
model granularity and (2) to adjust
midels appropriately
✔All
9.1
Evaluation of Optimisation vs.
Normalisation when designing or
implementing M & S capability
✔All
What level of modelling is required– when it
does it became a question of loss of value
rather than gain value (e.g. modelling
individual behaviour vs, group behaviour)
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
M & S Capability gaps 10/11/12/13
M & S Reqt
no
Sub M & S
Reqt NoDescription of Gap in M & S Capability Link to Dream Project Comments
10
Models of (1) human trust in CPS
devices/software and (2) software trust
in human co-worker expertise,
experience
✔Characterise and model dynamic
human interaction with CPS
✔Real-time system modelling
including human behaviour
✔Assisted Living
✔Educating CPS designers, operators,
maintainers, co-workers
11
Models of human decision-making, e.g.
case-based reasoning, natural decision-
making, and engineer-designed
protocols for decision-making –
constraint
✔Characterise and model dynamic
human interaction with CPS
✔Hybrid dynamic system verification
✔Real-time system modelling
including human behaviour
✔Assisted Living
No good approaches to address CPS systems
that learn – barrier. No good, run-time V&V
approaches available as a way to address the
learning issues
11.1Models of different human cognitive
styles
✔Hybrid dynamic system verification
✔Real-time system modelling
including human behaviour
What is going to be appropriate?
What attributes are necessary?
What differentiate a co-worker from the
other?
12
M & S to assess impact of different
training media and levels for CPS
operators and controllers
✔Characterise and model dynamic
human interaction with CPS
✔Real-time system modelling
including human behaviour
✔Educating CPS designers, operators,
maintainers, co-workers
It is about jobs and keeping manufacturing
Citizen science
How you prepare people to live in the CPS
world
12.1
Models of Human/intelligent agents
learning abilities and impacts -
single/double/triple loop learning.
✔Characterise and model dynamic
human interaction with CPS
✔Real-time system modelling
including human behaviour
✔Assisted Living
Individual agent learning abilities and impact
on roles and decision making
Same applies to groups of human/technical
agents
13Requirement for a platform to 'carry'
all the above M & S capability✔All
6. Workshop outputs: Expansion of relevant Dream Projects
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Area 5: Societal Aspects
During workshops on the previous 4 themes a set of Dream Projects had
emerged.
• Theme 1: Architectures Principles and models for Autonomous Safe
and secure Cyber-Physical Systems
• Theme 2: System Design, modelling and virtual engineering for
Cyber-Physical Systems
• Theme 3: Real time modelling for autonomous, adaptive and
cooperative Cyber-Physical Systems
• Theme 4: MBSE applied to Computing Platforms and energy
management
Each of these Dream Projects was considered in terms of its relevance to
Theme 5. The following set include a) 6 Dream Projects considered
relevant and hence embellished where necessary with issues that arose
during the Theme 5 Workshop [indicated in blue] and b) 2 new Dream
Projects: Assisted Living and Educating CPS designers, operators,
maintainers etc
Aim and
objectives of
Dream Project
Aim (of government): characterise and improve entry and use of CPS -> commons - public use of systems, public ownership
1. Gaining data about CPS in commons
2. Process of entry and use
3. Tools and methods – reasoning, assurance etc.
US contribution • Data, e.g.
• transportation (vehicle, security, incident, maintenance)
• security/defence (CBRNE, CT);
• lessons learned
• statute, policy, regulation, guidance
• successes, futures, opportunities for co-development
• taxonomy/ontology for management of CPS -> commons.
Desired EU
contribution
as US
Collaboration
instruments
• TTCP as template for exchanging security information
• virtual modelling environments
• scenario/context-based <- exemplars (cases, practices)
• design/architecture
• evaluation
• requirements (technical, social, political)
• statute, policy, regulation, guidance
• economic and technical integration, coherence (e.g. air traffic control)
Baseline state of
the art (key
existing
technologies)
Links to key
Theme 1
developments
in M&S
Other comments
Dream Project Characterise and Improve Entry and Use of CPSTeam
members:
• Individual, team standardised models for ‘human co-worker avatar’ do not exist = barrier
• No dynamic model of the ‘geographically-extended co-worker interface’ exists = barrier
• No meta-level tool exists to (1) decide required model granularity and (2) to adjust appropriately both models above = barrier
• No platform to carry all these tools exists = barrier• Need a ‘gamification’ approach to modelling to include
co-workers within CPS (i.e. CSR, RRI aspects) = barrier • Need models of (1) human trust in CP and (2) CPS trust
in co-worker expertise, experience - barrier
• See DARPA ‘New Social Science’ (NS2) Call, 2016• SoA model status is limited to 1-off partial, static models addressing usually 1, occasionally more, classes of performance; • No internal feed-back to update model parameters as model is operated.• No common, accepted ontology for modelling Co-worker behaviour
Aim and
objectives of
Dream Project
Aim: Verification of dynamic ensembles of hybrid systems in dynamic environmentsObjectives:• Modelling, simulation and experimentation frameworks• Formal specifications for desired objectives (security, resiliency, performance, etc)• Novel verification technologies and tools
US contribution • Autonomous systems research• Tool-chain’s for model-based design• Verification tools (model checking)
Desired EU
contribution
• Formal modelling and specification of desired behaviour• Interchange formats for hybrid dynamic systems• Verification tools (theorem provers)
Collaboration
instruments
• Joint workshops for framework exchanges• Development of joint tools (US/EU)
Baseline state of
the art (key
existing
technologies)
• Bounded model checker (dReal)• Theorem prover (Coq)• OpenMETA tool chain for model-based design• Frameworks for autonomous systems (CMU -> Uber, Google, etc)• Algorithms for emergent behaviour detection/avoidance• NSF ExCAPE project (Expeditions in Computing)
Links to key
Theme 2
developments
in M&S
• Scientific approaches to testing and evaluation of adaptive CPS• Hybrid modelling and simulation environments• Tools that capture key safety, security, resilience properties• Applications of new mathematical tools (e.g., networks)• Formal semantics frameworks for sound co-modelling• Complex system modelling for predicting emergent behaviour
Other comments
Dream Project Hybrid Dynamic System VerificationTeam
members:
• No need to include V&V for human co-workers/stakeholders, but if included, models of human performance are necessary.
• Include models of human decision-making• If CPS includes autonomous components (e.g.
humans, AI software), then V&V must address learning
• Must include models of human decision-making, e.g. case-based reasoning, natural decision-making, &engineer-designed protocols for decision-making =constraint
• No good approaches to address CPS systems that learn = barrier
Dream ProjectReal-time system modelling including human
behaviourTeam members: All
Aim, objectives &
application domain
of the project
Improve safety of CPS (including humans interactions) through improved models• Human interaction at several levels: design, interface, • Capturing the human’s intention – take into account interface • Authority sharing (human and AS) and mode confusion• Objective is to improve safety in unforeseen situations• Application domains automated transport, robots (e.g. care), • Focus is on individual interactions• Risk associated with use of bad models • Experimentation to populate models
(Desired) US
contributionBig data information to construct human model
(Desired) EU
contributionAs above
Collaboration
instruments.
Enablers & barriers
to collaboration.
Why should the EU
fund this project
(as opposed to an
alternative funder)?
Baseline state of
the art (key existing
technologies)
Related roadmap
elements
Developments in M&S: Collaboration opportunities:
• Individual and team ‘human co-worker avatar’ standardised models
• Abstract dynamic model of the ‘geographically-extended, co-worker interface’
• Meta-level tool to (1) decide required model granularity and (2) to adjust appropriately both models above
• Internal feed-back to update model parameters as model is operated.• Platform to carry all these tools• A ‘gamification’ approach to modelling to include responses
of/consequences for society and for co-workers within CPS (i.e. CSR, RRI aspects)
• Develop common, accepted ontology for modelling Co-worker behaviour • Models of (1) human trust in CPS devices/software and (2) software trust
in human co-worker expertise, experience - barrier• Blackbox
• See DARPA ‘New Social Science’ (NS2) Call• SoA is limited to 1-off partial, static models addressing usually 1, occasionally more, classes of performance; • No internal feed-back to update model parameters as model is operated.• No common, accepted ontology for modelling Co-worker behaviour • Must include models of human decision-making, e.g. case-based reasoning, natural decision-making, and engineer-designed protocols for
decision-making – constraint• No good approaches to address CPS systems that learn – barrier• Need models of (1) human trust in CPS devices/software and (2) software trust in human co-worker expertise, experience - barrier
Aim and objectives
of Dream Project
• Integrating and validating modelling/simulation tools in different domains (including data mining, human modelling, simulation of dynamical systems etc.)
• Establish a basis for comparison between different case studies (explore the benefits of different transportation system modes)
(Desired) US
contribution
• Data (testbeds, taxi companies, other private industry i.e. Google, etc.)
• Congestion models, simulation models
• Data, transportation network layout
(Desired) EU
contribution
As above
Collaboration
instruments /
enablers / barriers
• IP barriers, social acceptance
• Enablers: society demand, sustainability
• Enablers: technology maturity (vehicle autonomy, mapping / perception technology)
• Workshops, joint projects sponsored by both EU and US funding agencies, shared data respositories
Baseline state of
the art (key existing
technologies)
• Autonomous vehicle modelling / control
• System / fleet level control / modelling
• Mapping technologies
• Traffic modelling tools / simulations
Links to key Theme
x developments
in M&S
B: Predictive modelling & real-time decision support: • Real time decision support (M2M connection, HMI,
distributed and heterogeneous data)• Predictive modelling of different prod. cases and how
they effect quality, energy, environment, etc. 1: Architecture for fault identification2: First principles and empirical models: 1) humans 2) environment 3) perception 4) reasoning/decision making
3: Formal models of non-functional properties: safety, security, privacy, performance, cost, …5: Advanced learning algorithms with ability to train rapidly8: Big-data driven modelling9: High fidelity simulations that can be used for safety certification10: Hierarchical models for CPSE: Technology to assume trustworthy behavior by autonomous systems
Links to other
collaboration
opportunities
2: CPS verification & validation grand challenge – for adaptive learning CPS12: Integration of modelling approaches of various domains for autonomous systems with human involvement on a test case4: Collaborative use (or data sharing) of autonomous vehicle testbeds10: Integration of CPS models from different research communities (e.g. robotics + transportation)
Other comments
(inc. application
domain)
Transportation, robotics
Dream ProjectCase studies for autonomous transportation in EU/US
citiesTeam members:
• Mixed-mode transport studies (e.g. Munich)• UK legislation for autonomous vehicles• Consideration for autonomous vehicle users, other users of transport pathways• Recognition of role of transport pathway furniture• Access by users to information infrastructures for transport pathways
Aim and
objectives of
Dream Project
For a test problem that has a large Community of Practice, and many existing models, that has humans, legacy systems, legacy models, and current integration, find a way to move this to a higher level that is adaptive and resilient, and that can be shared. Must be done in a way that meets the desires of the society. The idea would be to use the project to evolve existing models to ones that can work together and continue to evolve into the future.
(Desired) US
contribution
• Collect existing models• Understand legal limitations• Develop and share requirements for the common model• Identify differences between states in the US
• Agree on a common interoperability standard• Apply standard on US systems• Evolve systems to new joint testbed
(Desired) EU
contribution
• Collect existing models• Understand legal limitations• Develop and share requirements for the common model• Identify difference between EU countries
• Agree on a common interoperability standard• Apply standard on EU systems• Evolve systems to new joint testbed
Collaboration
instruments
• Workshops to find the common problem• Workshops to develop standards (technical, interoperability)• Agreements for sharing (IP, technical data)• Funding structure – who contributes from each side• Joint demonstration (funding, logistics, location, preparation, etc.)
• Include standardization bodies from US and EU
Baseline state of
the art (key
existing
technologies)
• Existing legacy models• Existing legacy systems• Current integration approach• Machine intelligence• Protocols
• The approach could feasible for some domains but not for others
Links to key
Theme x
developments
in M&S
Model-based integration framework; a CPS testbed. Scientific approaches to testing and evaluation on adaptive resilient systemsAutomatic updates on virtual models when CPS components are movedHybrid M&S environmentsOnline model calibration
Links to other
collaboration
opportunities
• Dream project: federated EU/US test beds• It could be done in parallel with dream project on test beds
Other comments It will impact both sides on the Atlantic.Could be of interest to NATOProbably have to start with non-defense application to get support from both sides
Dream ProjectIntegration and Interoperability Models and Approaches
Team
members:
Aim and objectives
of Dream Project
• Collaborative analysis of data collected in the US & EU
• In order to study global trends in the Atlantic area
(Desired) US
contribution
• Focus on private transportation• More petrol fuelled cars• Rural area• Policy impact• Impact of heavy data processing – context aware
(Desired) EU
contribution
• Focus on public transportation• More diesel powered cars• Urban area / population density• Policy impact
Collaboration
instruments /
enablers / barriers
÷Funded / funding+ Strong research environments+ Personal / business relationship between US/EU researchers
Baseline state of
the art (key existing
technologies)
• Drone development very active• Machine learning / data collection / data
analysis• Environmental data capturing (sensors)• B1: Real time decision support (M2M
connection, HMI, distributed and heterogeneous data)
• B2: Predictive modelling of different prod. cases and how they effect quality, energy, environment, etc.
• A3: Technology to model product and system behavior in the context of changing environment
Links to key Theme
x developments
in M&S
• Autonomy• Real-time (location / time of data collection)• Adaptive - react to events
Links to other
collaboration
opportunities
• Different types of data (e.g. traffic)• 1: Case studies for autonomous
transportation in EU/US cities
• 4: Collaborative use (or data sharing) of autonomous vehicle testbeds
• 13: Collaborative development between geographically dispersed teams (24-hour development)
Other comments
(inc. application
domain)
Dream Project Autonomous collection of env. data via drones Team members:
• Since environmental data most likely will also include human data, privacy issues will be important –privacy, protection, safety, security
Template for Dream Collaborative ProjectsAssisted Living
The aim is to investigate control and safety issues involved in a socio-technical CPS [ie comprising autonomous devices as well as intelligent
human and software components. Within this issues to be investigated include: absolute need for safety and privacy in relation to
vulnerable people; how do healthcare devices and systems adapt to the ‘vanishing mind’; ethics and trust throughout the CPS; engineering
governance; ownership of CPS etc. Identify limitations and develop new M & S capability for the above & particularly the human – system
interface
SoA in use of robots/autonomous systems in Healthcare; modelling of financial provision and legal issues for private healthcare; models for
health care system optimisation, models of autonomous systems’ ability to interpret non verbal communication; control systems
simulations; simulation capability for predictive human performance modelling; models of dynamic functional reallocation and authority
As for US but modelling should focus on publicly financed Systems
• Workshops to find the common problem
• Workshops to develop standards (technical, interoperability)
• Agreements for sharing (IP, technical data)
• Funding structure – who contributes from each side
• Joint demonstration (funding, logistics, location, preparation, etc.)
Because the EU has a rapidly increasing aged population with health and social care needs. Current systems are overloaded and inefficient
and more efficient systems are required. Autonomous systems are the future but there are increasing concerns in the public about trust in
/safety and capability of such systems
Also not only safety but acceptance of these systems, how can a robot or technology be empathic or at least act as such, or at least be
perceived as such through interaction, presence and look (dog shape versus spider shape)
o Autonomous systems are the future but in the Healthcare domain they are used in specialised areas such at hospital theatres. Use in
wider health and social care systems is at a very low TRL level. Testing and simulation of various scenarios is not available at the
moment and situation awareness models are also lacking across such systems
Template for Dream Collaborative ProjectsEducating CPS designers, operators,
maintainers, co-workers
The aim is to address the current skill gaps that exist in both the US and EU at all levels in terms of development, operation and maintenance
of CPS systems. There is also a need, having identified the gaps, to design and provide multi-dimensional education provision for CPS
Systems
Statement of requirements at all levels eg higher Education, college, apprenticeships etc; Gap analysis of provision against need; M& S
training simulation capability; accreditation modelling to establish creditability and to encourage movement of trained personnel across the
US and different application domains
As for US but modelling should focus on education provision at Higher Ed, Further Ed, Apprenticeships. NVQs etc and also on movement
across different countries in the EU
• Workshops to find common problems and differences. Also to establish a set of requirements for trained personnel
• Sharing of expertise and relevant content at the different educational levels
• Funding structure – who contributes from each side
• Joint demonstration (funding, logistics, location, preparation, etc.)
Because it is accepted that CPS systems are pervasive and increasing and do not recognise domain or national borders. It is also accepted
that a skills gap exists and a range of appropriate training options and delivery mechanisms need to be developed and applied according to
cost time and quality variables.
Not known exactly at this time as not aware of any in-depth analysis of current provision. However all major organisations eg Artemis and
many EU and US funded projects have identified the provision of trained people and a key factor in the successful deployment of CPS across
all domains
7. Workshop outputs: Key Discussion points arising
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
US View on CPS
• CPS implementation
o Needs legislative framework to be implemented
• Definitions of CPS:
o The US definition of CPS presented by TAMS4CPS is just
one of the US definitions of CPS: there are many other
points of view in US on what is a CPS.
o The US CPS focus started with the NSF call (Finance
domain), then people started to call everything CPS (e.g.
research in embedded system was relabelled CPS)
o It was much later in US that human elements were
included or considered as being into CPS
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Issues and Trends
• Autonomy and automation
o Autonomy it is an important aspect especially in the military domain
o Automation is a sensitive issue particularly related to the impact on jobs
and employment
o Lack of predictive models for the impact of both - approach was mainly
to go ahead and do it and then see what it happens
• The importance of the East (e.g. China) in CPS models because parts of CPS
are and will continue to be produced in China
• CPS and Cloud 2 should be closely related
• Cross-disciplinary integration (US is of the opinion that EU does it better)
• Assisting living is domain of interest for the US
• The data transfer issue between human and machine (the pipe is small)
o The response time is different for human and machine
TAMS4CPS - Trans-Atlantic Modelling and Simulation for Cyber-Physical Systems
Issues and Trends
• Robots
o Legal aspects/ethics
o Intelligence (within design)
o Testing and simulation of various scenarios is not available at the moment
o The co-worker situational awareness issue
• The problem of “updates” of various systems within the CPS (the updates don’t
happen necessarily at the same time)
• Gaps in terms of modelling capability:
o Modelling at different levels/layers. Models at ‘organisation’ level are
difficult although methods exist to model the layers underneath
o At the moment the majority of models are at a semantic level
o At the moment the majority of models are static – lack of dynamic models
or continuous predictive simulations
o Optimisation vs. normalisation (link with value); level of modelling – where
it becomes a question of loss of value rather than gain value (e.g. modelling
individual behaviour vs, group behaviour)
8. Conclusions and Next Steps
Conclusions: outline
• There is a dearth of M & S capability dealing with societal aspect of socio-technical CPS
• There is a need for cross-disciplinary teams to work together on the M & S capability gaps identified
• There is a need to develop M&S capability for larger scale dynamic CPS simulations embracing communities
• Similar issues exist in US and EU with some differences of interpretation: the rise of autonomy in CPS is a high common factor
Next steps
• This report will be reviewed by workshop participants and will then be published online on the project website (www.tams4cps.eu).
• Material from this report (and those for the other themes) will be used to develop formal deliverables later this year:
• State of the art: October 2016
• Test cases: October 2016
• Strategic Research Agenda for Collaboration: September 2016 (draft) & December 2016 (final)
Workshop Dates Location
Theme 1 US
workshop
8-10 July, 2015 George Mason University,
Washington D.C.
Theme 2 US
workshop
12-14 October,
2015
Georgia Tech, Atlanta
Theme 1 + 2
EU workshop
9-10 November,
2015
Brussels, Belgium
Theme 3 US
workshop
10-11
December, 2015
Purdue University,
Lafayette
Theme 3 EU
workshop
11-12 February,
2016
Brussels, Belgium
Theme 4 US
workshop
17-18 March,
2016
UTSA, San Antonio
Theme 5 US
workshop
16-17 May,
2016
Stevens Institute,
Hoboken
Theme 4 + 5
EU workshop
(in parallel)
16-17 June,
2016
Kongsberg, Norway
9. Acknowledgements
Acknowledgements
• The consortium is grateful for the support of our US host Prof Greg Vesonder and the workshop participants. The workshop would not have been a success without his active involvement.
Top Related