Tufts Wireless Laboratory Tufts University School Of Engineering Cyber-Physical Systems: A New Frontier Nicole Ng 10/19/09
Tufts Wireless Laboratory Tufts University School Of Engineering Why Cyber-Physical Systems are needed • Major challenges today: – Global warming and energy shortage – Rapid aging of fraction of world’s population and chronic diseases that threaten to bankrupt healthcare services
Tufts Wireless Laboratory Tufts University School Of Engineering Why Cyber-Physical Systems are needed • Global Warming and Energy Shortage – Building and transportation have heavy energy consumption • Buildings – Goal to have Net Zero Energy buildings where 60 -70% efficiency gains required for reducing demand balance is supplied by renewable energy – Need CPS to manage flows of mass, energy, and information in a coherent way • Transportation – Work related travel is a significant fraction of transportation cost – Telepresence research => goal to make all interactions seem local rather than remote – Need to integrate wired and wireless networks with real-time, interactive, immersive three-dimensional environments and tele-operation
Tufts Wireless Laboratory Tufts University School Of Engineering Why Cyber-Physical Systems are needed • Rapid aging of fraction of world’s population and chronic diseases – Chronic diseases account for 70% of all deaths in the U. S – Advanced biotechnology has potential to improve health of aging population • Stem-cell biotechnology – stem cells directed to differentiate into specific types => renewable source of replacement cells and tissues used to treat diseases or test new drugs • Need to mimic complex and dynamic stem-cell microenvironment – Independence • Major cost is loss of ability to remain at home because of need for greater health care supervision • Assistance in physical mobility, cognitive impairment • CPS => monitor health remotely
Tufts Wireless Laboratory Tufts University School Of Engineering Challenges of CPS • • • Real-time system abstractions Robustness, safety and security of CPS System Qo. S Composition Challenge Systems engineering research Trust in CPS
Tufts Wireless Laboratory Tufts University School Of Engineering Real-time system abstractions • Existing hardware design and programming abstractions for computing => principle task is data transformation, but CPS are real-time systems • Need to re examine existing hardware and software architectures, need to re think split between programming language and operating systems • In CPS, computation interacts with physical world, need to deal with events in time and space • Need strong real-time concurrent programming abstractions
Tufts Wireless Laboratory Tufts University School Of Engineering Real-time system abstractions • Programming abstractions need support at middleware and operating system layers for: – Real-time event triggers – Consistent views of distributed states in real-time within sphere of influence – Topology control and “dynamic real-time groups” in form of packaged service classes of bounded delay, jitter and loss under precisely specified conditions – Interface to access same type of controls regardless of underlying network technology
Tufts Wireless Laboratory Tufts University School Of Engineering Robustness, Safety and Security • Goal to capture uncertainty, errors, faults, failures, and security attacks • Regard most errors as transient or permanent • Intermittent errors pose a problem – Last several milliseconds to a few seconds – Need to apply algorithms that do not rely on accuracy of one computation • CPS architecture safety critical services must be guaranteed by a small subset of modules and their interactions – Assumptions about physical environments should be fully tested – Need to develop advanced integrated static analysis and testing technologies to ensure that: • 1) software code is compliant with design and • 2) assumptions about external environment are sound
Tufts Wireless Laboratory Tufts University School Of Engineering Robustness, Safety and Security • CPS are deeply embedded, will evolve in place • Need to handle known and unknown residual errors and security gaps • Loops must be closed across cyber and physical • System must be able to effectively counter-act uncertainties, faults, failures, and attacks
Tufts Wireless Laboratory Tufts University School Of Engineering System Qo. S Composition Challenge • CPS are distributed and hybrid real-time dynamic systems • New theory of system composition must provide comprehensive treatment of system integration concerns – System composition: Qo. S properties and functional correctness of system can be derived from architectural structure, subsystem interaction protocols, and local Qo. S properties and functional properties of various constituent components • CPS system properties depend on both component properties and the structure of the system architecture • Must address composability at each Qo. S dimension and how protocols interact
Tufts Wireless Laboratory Tufts University School Of Engineering Systems Engineering Research • Need combination of model-based system and software design and integration technologies, and deep analysis of underlying abstractions and their interactions • Challenge: formulate new calculus that merges time-triggered and event-driven systems
Tufts Wireless Laboratory Tufts University School Of Engineering Trust in CPS • Combination of reliability, safety, security, privacy, and usability • Need to incorporate fault models • Create guards to ensure that automation does not increase hazards compared to non-automated system • 2 cyber physical systems must never interact directly • Privacy and security concerns – Personal data and medical data should be protected with different levels of information disclosure – Need privacy protection • Usability – Human cognition, computer-human interaction, and interface design
Tufts Wireless Laboratory Tufts University School Of Engineering Medical Device Network: An example CPS system • Next generation medical system: ubiquitous system of wired and wireless networked medical devices and medical information systems for secured, reliable, privacy-preserving health care • Networked system that improves quality of life • Ex) Surgical operation – Context information (ex. sensitivity to drugs) automatically routed to relevant devices (ed. Infusion pumps) – Changes in vital signs correlated with streams of imaging data – Expert surgeon can remotely carry out operation using remote displays and robot-assisted surgical machines while in another country – Challenges: • 1) Different devices and subnets have different levels of clinical criticality • 2) Moving wired networks to wireless networks
Tufts Wireless Laboratory Tufts University School Of Engineering Reference “Cyber-Physical Systems: A New Frontier” Lui Sha, Sathish Gopalakrishnan, Xue Liu, and Qixin Wang University of Illinois at Urbana Champaign, University of British Columbia, Mc. Gill University 2008 IEEE International Conference on Sensor Networks, Ubiquitous, and Trustworthy Computing
Скачать презентацию Tufts Wireless Laboratory Tufts University School Of Engineering
4383f59e5be58475668b9060b5120c12.ppt