TY - GEN
T1 - Time in cyber-physical systems
AU - Shrivastava, Aviral
AU - Derler, Patricia
AU - Baboud, Ya Shian Li
AU - Stanton, Kevin
AU - Khayatian, Mohammad
AU - Andrade, Hugo A.
AU - Weiss, Marc
AU - Eidson, John
AU - Chandhoke, Sundeep
N1 - Funding Information:
National Science Foundation grants CNS 1525855, and NIST grant 60NANB15D322.
Publisher Copyright:
© 2016 ACM.
PY - 2016/10/1
Y1 - 2016/10/1
N2 - Many modern cyber-physical systems (CPS), especially industrial automation systems, require the actions of multiple computational systems to be performed at much higher rates and more tightly synchronized than is possible with ad hoc designs. Time is the common entity that computing and physical systems in CPS share, and correct interfacing of that is essential to flawless functionality of a CPS. Fundamental research is needed on ways to synchronize clocks of computing systems to a high degree, and on design methods that enable building blocks of CPS to perform actions at specified times. To realize the potential of CPS in the coming decades, suitable ways to specify distributed CPS applications are needed, including their timing requirements, ways to specify the timing of the CPS components (e.g. sensors, actuators, computing platform), timing analysis to determine if the application design is possible using the components, confident top-down design methodologies that can ensure that the system meets its timing requirements, and ways and methodologies to test and verify that the system meets the timing requirements. Furthermore, strategies for securing timing need to be carefully considered at every CPS design stage and not simply added on. This paper exposes these challenges of CPS development, points out limitations of previous approaches, and provides some research directions towards solving these challenges.
AB - Many modern cyber-physical systems (CPS), especially industrial automation systems, require the actions of multiple computational systems to be performed at much higher rates and more tightly synchronized than is possible with ad hoc designs. Time is the common entity that computing and physical systems in CPS share, and correct interfacing of that is essential to flawless functionality of a CPS. Fundamental research is needed on ways to synchronize clocks of computing systems to a high degree, and on design methods that enable building blocks of CPS to perform actions at specified times. To realize the potential of CPS in the coming decades, suitable ways to specify distributed CPS applications are needed, including their timing requirements, ways to specify the timing of the CPS components (e.g. sensors, actuators, computing platform), timing analysis to determine if the application design is possible using the components, confident top-down design methodologies that can ensure that the system meets its timing requirements, and ways and methodologies to test and verify that the system meets the timing requirements. Furthermore, strategies for securing timing need to be carefully considered at every CPS design stage and not simply added on. This paper exposes these challenges of CPS development, points out limitations of previous approaches, and provides some research directions towards solving these challenges.
UR - http://www.scopus.com/inward/record.url?scp=84995520569&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84995520569&partnerID=8YFLogxK
U2 - 10.1145/2968456.2974012
DO - 10.1145/2968456.2974012
M3 - Conference contribution
AN - SCOPUS:84995520569
T3 - Proceedings of the 11th IEEE/ACM/IFIP International Conference on Hardware/Software Codesign and System Synthesis, CODES 2016
BT - Proceedings of the 11th IEEE/ACM/IFIP International Conference on Hardware/Software Codesign and System Synthesis, CODES 2016
PB - Association for Computing Machinery, Inc
T2 - 11th IEEE/ACM/IFIP International Conference on Hardware/Software Codesign and System Synthesis, CODES 2016
Y2 - 1 October 2016 through 7 October 2016
ER -