Cyber-Physical Systems: Challenges for Systems Engineering agendacps Closing Event April 12th, 2012, EIT ICT Labs, Berlin Eva Geisberger fortiss An-Institut der Technischen Universität München
Cyber-Physical Systems: Scenarios Dr. Eva Geisberger CPS: Challenges for Systems Engineering, agendacps Closing Event, April 12th, 2012 2
Evolution towards CPS by means of the brake functionality 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 - diversity of use risks basic physical functionality multifunctional embedded system logical causal chains active steering and controlling monofunctional stability is supported embedded system while braking reactive steeering and control manoeuvrability is preserved while braking environment multifunctional system of systems simple physical context model active steering and autonomous control automatic braking in critical situations environment / domain model adaptive and interactive action control systems cooperation in open social contexts complex context and domain models comprehensive distributed data analysis for braking support also braking remotecontrolled by infrastructure (e.g. by police and first responder) 1978 1995 2003 (upcoming) time Dr. Eva Geisberger CPS: Challenges for Systems Engineering, agendacps Closing Event, April 12th, 2012 3
CPS imply interactive evolution to open, adaptable and cooperative socio-technical systems application domain X e.g. health application domain Y e.g. mobility area 2 area 1 area 3 area 1 area 2 area 3 operating across multiple domains in different levels of uncertain physical and social environments: controlled specified uncertain Dr. Eva Geisberger CPS: Challenges for Systems Engineering, agendacps Closing Event, April 12th, 2012 4
CPS capabilities CPS are required to be x-aware and assimilable to their physical/social context (correct perception and interpretation of) situation and context, self, third party, and human (state, objectives, intention, ability to act), capable of learning and adaptable, transparent, equipped with predictable human machine interaction, reliable, cooperative, capable to decide and act strategically, subject to risk, target and quality analysis (quality in use, quality of service) as well as QoS assurance. Dr. Eva Geisberger CPS: Challenges for Systems Engineering, agendacps Closing Event, April 12th, 2012 5
CPS detailed capabilities Dr. Eva Geisberger CPS: Challenges for Systems Engineering, agendacps Closing Event, April 12th, 2012 6
Beside the technological challenges to realize the CPS capabilities: multi-criteria situation assessment, cooperation and negotiation with multi-agent systems, intention and plan recognition, user modeling, machine learning, data mining, evolution, self-organization, etc. basic technologies: domain modeling, sensor and actuator technology, efficient and reliable communication infrastructure, platform and parallel processing units, distributed, stable real-time controlling, etc. building up, adapting and managing cross-domain CPS infrastructures, platforms and standards with basic services for interoperability (technical, semantically, user-perceived) and quality of service (dependable, secure, efficient), etc. the major success factors and challenges of CPS are subject to users and social acceptance of CPS and their experienced quality of interaction or expected behavior & control in every day scenarios. Dr. Eva Geisberger CPS: Challenges for Systems Engineering, agendacps Closing Event, April 12th, 2012 7
Challenges for systems engineering Human-system cooperation, usability, controllability and safety: shared control, transparency/controllability, integrated models for human-machine interaction, cooperative and strategic task and action modeling, conflict and reconciliation handling Formal and integrated system and architecture models: Integrated requirements, context, domain, system modeling Hybrid system and architecture: analog-digital control models, human-machine, socio-technical interaction and network models Interoperable reference architectures, domains and platforms Expanded quality modeling and engineering standards: Models for quality in use, quality of service, compliance, technical and organizational models and methods for quality assurance Elicitation and negotiation of acceptance requirements and corresponding system concepts (e.g. governance, fairness, etc.) Dr. Eva Geisberger CPS: Challenges for Systems Engineering, agendacps Closing Event, April 12th, 2012 8
Social challenges imply new engineering strategies 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 behavior self-determined usable and controllable by the users compatible with non-networked systems and services (as well as dropouts) Inter-, transdisciplinary and explorative research and development, which require integrated models and methods of virtual engineering Interactive innovations by means of economic ecosystems and platforms, regional & international innovation systems integrate different life-cycles, business models and engineering cultures of CPS-components and services complementary concepts of competition, cooperation and distributes value creation Dr. Eva Geisberger CPS: Challenges for Systems Engineering, agendacps Closing Event, April 12th, 2012