[PPT] - Cyber-Physical Systems: imminent challenges Mar a Victoria PowerPoint Presentation

SLIDE 1

Cyber-Physical Systems: imminent challenges

Mar ´ ıa Victoria Cengarle

fortiss joint work with Manfred Broy Eva Geisberger

Tech. Univ. Munich

fortiss

March 20th, 2012

Monterey Workshop ’12

SLIDE 2

Project goal The project aims at capturing technology trends and innovation po- tential in the context of Cyber-Physical Systems (CPS) in a holistic and systematic fashion, and at the derivation of implications for key research and action areas. By means of crucial application fields the significance is illustrated that CPS have for economy and society. Ul- timate goal of the project is to strengthen and expand Germany’s position in the field of CPS.

M. V. Cengarle

2 Monterey Workshop ’12

SLIDE 3

Project plan

Goals / Requirements Capabilities Scenarios Characteristics Technologies Implications Challenges

M. V. Cengarle

3 Monterey Workshop ’12

SLIDE 4

Scenario smart mobility

M. V. Cengarle

4 Monterey Workshop ’12

SLIDE 5

Scenario smart mobility (contd)

M. V. Cengarle

5 Monterey Workshop ’12

SLIDE 6

Scenario smart health

M. V. Cengarle

6 Monterey Workshop ’12

SLIDE 7

Scenario smart health (contd)

M. V. Cengarle

7 Monterey Workshop ’12

SLIDE 8

Scenarios smart mobility and smart health

M. V. Cengarle

8 Monterey Workshop ’12

SLIDE 9

CPS domain structure

M. V. Cengarle

9 Monterey Workshop ’12

SLIDE 10

CPS classification According to social and spatial network structures (topologies): (1) large-scale infrastructure systems and services (controlled area) (2) social infrastructure systems and services (defined area) (3) social application systems and services – including businesses; demarcated area, spanning over diverse domains,

pen to individual application systems and utilisation processes
M. V. Cengarle

10 Monterey Workshop ’12

SLIDE 11

CPS key drivers Smart embedded systems, mobile services, ubiquitous computing Internet as business web with two complementary forms: – transfer to the “cloud” – networked components (e.g., RFID technology) Semantic Web, techniques of Web 2.0, interactive design of integrated services through – user-defined interaction, configuration of knowledge, integrated services – communities of developers

M. V. Cengarle

11 Monterey Workshop ’12

SLIDE 12

Evolution towards CPS by means of the brake

1978

basic physical monofunctional adaptive and interactive multifunctional

functionality time

funtionality systems’ cooperation action control in open social context human−system cooperation complex context and domain model comprehensive distributed data analysis for braking support system of systems also braking remote−controlled by infrastructure (e.g. by police embedded system embedded system while braking is preserved manoeuvrability and controlling reactive steering logical causal chains and controlling active steering while braking stability is supported context model simple physical active steering and autonomous control automatic braking in critical situations multifunctional

With each stage of evolution increased − user−centric functionality − networking and integration with the context − complexity of possible causal chains − coordination of systems involved − human−machine cooperation

context

context & domain model

HMI HMI

and first responder)

− diversity of use risks 1995 2003 (upcoming)

M. V. Cengarle

12 Monterey Workshop ’12

SLIDE 13

CPS characterisation By CPS, physical and virtual, locally/globally networked systems are fused into systems of systems with dynamically shifting boundaries that are context-adaptive, partially or completely autonomous, and capable of active real-time control, cooperative with each other under distributed, alternating control, and able of comprehensive human-system cooperation.

M. V. Cengarle

13 Monterey Workshop ’12

SLIDE 14

CPS characterisation (contd) CPS include embedded systems, logistics, coordination and manage- ment processes as well as Internet services that, using sensors, directly capture physical data and, through actuators, act on physical processes that are interconnected by means of digital networks, use globally avail- able data and services, and have multimodal human-system interfaces. CPS are open socio-technical systems that provide a range of new functionalities, services and features which go far beyond the current capabilities of embedded systems with controlled behaviour.

M. V. Cengarle

14 Monterey Workshop ’12

SLIDE 15

CPS capabilities CPS are required to be

x-aware and assimilable to their physical/social context,
capable of learning and adaptable,
transparent, equipped with predictable human-machine interaction,
reliable, cooperative, strategic,
subject to risk, target and quality analysis as well as QoS assurance.
M. V. Cengarle

15 Monterey Workshop ’12

SLIDE 16

CPS detailed capabilities

(1) Cyber-physical sensor and actuators technology, virtual, locally/ globally networked, with real-time management (2) Systems of systems (SOS), controlled network with dynamic boundaries (3) Context-adaptive and (partially) autonomous systems (4) Cooperative systems with distributed, alternating control (5) Comprehensive human- system cooperation Central abilities and non- functional requirements, quality in use, quality of service (QoS)

 Parallel acquisition (through

sensors), fusion, processing of physical data from the local/ global environment in real time (physical awareness)

 Interpretation of the situation

w.r.t. the goal achievement and job completion of the CPS

 Acquisition, interpretation,

deduction, prediction of faults,

bstacles, risks

 Interaction, integration, rules

and control of CPS components and functions

 Globally distributed, networked

real-time control

 Interpretation of data from

context and situation over several levels, depending on application situations

 Targeted selection,

incorporation, coordination and use of services— depending on situation, local and global

bjective, and behaviour

 Service composition and

integration, decentralised controls: recognition of missing services, data and functions, and active search and dynamic incorporation of them

 Evaluation of components and

services to be incorporated regarding use and quality required for the application (QoS, overall quality) as well as possible risks

 Reliability and compliance

w.r.t. guaranteed QoS

 Controlled access to system’s

wn data and services

 Extensive, continuous context

awareness

 Continual collection,

bservation, selection,

processing, evaluation, communication of context data, situation data and application data (often in real time)

 Targeted adaptation of the

interaction, coordination, control with/ of other systems and services

 Recognition, analysis and

interpretation of plans and intentions of systems and participating users

 M odel creation for application

field, application domain, available services, tasks, and participants incl. their roles,

bjectives and requirements

 Assessment of objectives and

steps to achieve them, taking into consideration alternatives concerning costs and risks

 Self-awareness in terms of

knowledge about own situation, status and options for action

 Learning of e.g. modified work

processes, logistics, habits, interaction, etc., and corresponding behaviour adaption

 Self-organisation  Distributed, cooperative and

interactive perception and evaluation of the situation

 Distributed, cooperative and

interactive determination of the steps to be carried out— depending on the evaluation of the situation, of the objectives

f individual participants and of

the objectives of the community including these participants (local vs. global

bjectives)

 In doing so, coordinated

estimation and negotiation of the decision ultimately taken (i.e. own and shared control and decision-making autonomy)

 Decision with uncertain

knowledge

 Cooperative learning and

adaptation to situations and needs

 Estimation of the quality of

wn and external services and

abilities

 Coordinated processing of mass

data

 Intuitive, multimodal, active

and passive HM I support (simplified control)

 Support of a further (time and

space) and enlarged perception, capacity to act for individual and several persons (groups)

 Recognition and interpretation

f human behaviour including

feelings, needs and intentions

 Acquisition and evaluation of

state and context of human and system (extension of perception and of evaluation skills)

 Integrated and interactive

decision making and action of systems and individual persons

r multitudes

 Ability to learn

Required capabilities

 “ X” awareness (correct

perception and interpretation

f)
situation and context
self, third party, and human

(state, objectives, intentions, ability to act)

 Learning and adaption

(behaviour)

 Self-organisation  Cooperation, negotiation and

decision-making (within precise boundaries—compliance)

 Decisions with uncertain

knowledge

 Policy-making and, if

applicable, compliance with QoS guarantees

 Comprehensive safety and

security policies

 Transparent HM I, shared

control & integrated situation evaluation and predictable action

 Risk management  Proactive, strategic and reliable

action

 Privacy protection  increasing openness, complexity, autonomy, “ smartness” and evolution of the systems (with disruptive effects in the fields of application) 

M. V. Cengarle

16 Monterey Workshop ’12

SLIDE 17

Technical challenges Infrastructure and standards Openness spanning over diverse domains Formal & integrable requirements, context and domain modelling, hybridism; component and reference architectures Situation and self awareness, adaption and evolution Shared control, conflict and reconciliation handling Human-system cooperation Safety and security, personal data protection

M. V. Cengarle

17 Monterey Workshop ’12

SLIDE 18

Technologies Physical situation recognition: sensor fusion, pattern recognition, situational map Planning & anticipatory, partially or completely autonomous behaviour: multi-criterial situation assessment, artificial intelligence Cooperation and negotiation: multi-agent systems Human-machine interaction: human-machine interface and interaction modalities, intention and plan recognition, user modelling, human awareness Learning: machine learning and data mining Evolution, strategies of self organisation and adaption: self organising manufacture, self organising communication networks

M. V. Cengarle

18 Monterey Workshop ’12

SLIDE 19

Basic technologies Domain modelling, ontologies and domain-specific languages Sensor and actuator technology Communication infrastructure and plattform Efficient parallel processing units Distributed stable controlling

M. V. Cengarle

19 Monterey Workshop ’12

SLIDE 20

Social challenges Acceptance, which calls for participatory analysis and design of – systems and services that are manageable, tailorable, trustworthy, fault tolerant, accountable – capable of learning from user’s behaviour – ergonomic (cf. human factors) – compatible with non-networked systems and services (as well as dropouts) Inter- and transdisciplinary research and development Economic ecosystems Social guidelines and stipulations

M. V. Cengarle

20 Monterey Workshop ’12

SLIDE 21

Implications Great opportunity for academia, industry, business and economy: – Challenges for technology – Interdisciplinarity, value creation, innovation – Changes in business, law, and politics Contribution to accident prevention, smart use of limited resources, etc.

M. V. Cengarle

21 Monterey Workshop ’12

SLIDE 22

Implications (contd) (Part of) Sixth wave of innovation

M. V. Cengarle

22 Monterey Workshop ’12

SLIDE 23

Download area Position paper: www.fortiss.org/fileadmin/user_upload/downloads/ agendaCPS_Position-paper.pdf Survey (in German): www.fortiss.org/fileadmin/user_upload/downloads/ agendaCPS_Studie.pdf

M. V. Cengarle