Visible to the public 2012 National Workshop on the New Clockwork for Time-Critical Cyber-Physical Systems

Visible to the public 

Submitted by Katie Dey on Mon, 10/01/2012 - 10:16pm

The NITRD National Workshop on The New Clockwork for Time-Critical Cyber-Physical Systems workshop was held on October 25-27, 2012 at the Hyatt Regency in Baltimore, Maryland. The general session was on Thursday, October 25 and Friday, October 26 with a program committee report writing session on Saturday, October 27.

Workshop presentations can be found by clicking on the files tab at  the left menu bar. The workshop report is in progress. For further information about the workshop and the workshop report, please contact Raj Rajkumar at raj@ece.cmu.edu.

OBJECTIVE AND BACKGROUND

The High Confidence Software and Systems (HCSS) Coordinating Group (CG) of the Networking and Information Technology Research and Development (NITRD) subcommittee sought input from researchers, industry, and government agencies regarding problems, questions, and new directions for research on technology for time-critical systems. Each new generation of cyber-physical systems (CPS) raises the level of trust that people must put in these systems, while achieving assurances that the systems are worthy of that trust becomes more difficult. Examples of critical reliance on cyber-physical systems can be seen in defense systems, civil aviation, highways, energy production, advanced manufacturing, and modern healthcare. These systems make increasingly complex demands for real-time coordination among distributed subsystems. Even the current generation of large-scale real-time cyber-physical systems may unpredictably miss timing requirements, and expose the whole system to the risk of failure. Certainly progress on future systems will suffer without a better theory and practice of timed CPS. The solution will require contributions from researchers and practitioners from all relevant fields including control, embedded systems, hardware, networking, real-time systems, security, sensing, software, and timing.

Time has always been a critical issue in science and technology. Time measurement, distribution, and agreement technology has reached an important inflection point providing much more accurate time references on physical scales ranging from microsystems to global systems. Advances in distributed clock synchronization technology, such as GPS time and IEEE 1588, create new opportunities and challenges. On one hand, widely distributed systems can have a nearly synchronous view of the current time, thereby enabling better functionality and greater reliability so long as the synchronization mechanism works. On the other hand, mechanisms for time synchronization do fail. Natural phenomena and intentionally malicious attacks can disrupt timing in complex systems with catastrophic effects. Several recent trends have exacerbated time-related problems by increasing reliance on technologies with intrinsic time variability such as wireless communication, multi-core processors, virtualization, and cloud computing. Advances in the science of time-critical systems are needed if we are to be able to exploit the benefits of these technologies with assurance that systems will operate safely and reliably. We call these advances the “new clockwork”.

The goal of the workshop was to define a list of needs for research on time-critical aspects of cyber-physical systems so that future research can develop robust foundations for reasoning about time in cyber-physical systems across scales, managing resources to meet timeliness requirements, and ensuring service agreements through new tools, techniques and methodologies. The workshop was structured as a sequence of panels, presentations and breakout sessions. The workshop produce a report for the HCSS agencies.

TOPICS
Workshop topics included but are not limited to:

The new clockwork and clock synchronization

  • High confidence time reference (availability, accuracy, traceability)
  • Modern IC designs, system clocks, multiple time domains, jitter, PLLs, power considerations, CSACS
  • Impact of synchronized clocks on OS, hardware interfaces…
  • Security of clock synchronization
  • Systems of systems and hierarchical clock synchronization
  • Standards, protocols, and technologies for clock synchronization (current and missing)
  • Mobility issues in clock synchronization
  • Formal models of time
  • Next generation time services

Application domains, challenges and algorithms

  • New application areas and algorithms enabled by the new clockwork
  • Temporal semantics of data, control and monitoring
  • Beyond periodicity in sensing actuation and control
  • New clockwork impact on performance
  • Energy considerations and opportunities enabled by the new clockwork
  • Wireless networks and timing
  • Open source experimental platforms

Time criticality

  • Timing robustness, determinism, predictability
  • Management and scheduling of physical, computational, and communication resources
  • Error detection and management in the new clockwork
  • Performance, and temporal issues related to distributed systems
  • Mobility issues
  • Continued operation in the presence of degraded or changing synchronization
  • Mixed criticality and mixed timing sensitivities

Verification, validation, vulnerability assessment and security

  • Assurance and certification of time critical systems e.g. layered assurance
  • Resilience in time critical systems (heterogeneous, fault tolerance, multiple sources)
  • Formal models of event timing and timed systems
  • Potential impact and new vulnerabilities of timing on security
  • Use of time to improve security
  • Privacy issues in systems based on the new clockwork
  • Provenance in systems based on the new clockwork

Architecture, systems and tools

  • System and component architectures for the new clockwork
  • Design space from event-based to time-based systems (including combinations)
  • Open, adaptive, evolvable and ad hoc systems
  • Cooperative multi-systems
  • Multicore and multiprocessor on chip timing issues
  • Time semantics in languages, Oss, and hardware, in simulation, execution
  • Ontology of time
  • Missing standards, protocols, and technologies to enable all of this
  • Models of timed systems: e.g. synchronized vs. synchronous systems
  • Tools and methodologies for the complete system design and implementation cycle

 

GENERAL CO-CHAIRS

  • Edward A. Lee, UC Berkeley
  • Raj Rajkumar, Carnegie Mellon

PROGRAM COMMITTEE

  • John Eidson, UC Berkeley
  • Raj Rajkumar, Carnegie Mellon
  • John Rushby, SRI International
  • Aloysius K. Mok, University of Texas

 

For more information about the workshop, please contact the workshop organizers at raj@ece.cmu.edu or eidson@eecs.berkeley.edu.
 

This event is funded by the National Science Foundation (NSF) and the National Security Agency (NSA). Appropriate acknowledgement of this support should be included in reports or publications based on work performed under this award.