TY - GEN
T1 - Model-based availability analysis for automated production systems
T2 - 15th ACM-IEEE International Conference on Formal Methods and Models for System Design, MEMOCODE 2017
AU - Mund, Jakob
AU - Junker, Maximilian
AU - Bougouffa, Safa
AU - Cha, Suhyun
AU - Vogel-Heuser, Birgit
N1 - Publisher Copyright:
© 2017 Copyright is held by the owner/author(s).
PY - 2017/9/29
Y1 - 2017/9/29
N2 - Availability is among the most important characteristics of manufacturing systems since it affects the productivity of the system. Yet, we argue that current approaches are inadequate for thoroughly analyzing availability of such systems. In this paper, we present a novel approach for analyzing availability of automated production systems. This approach extends the system's normative specification with deviation models to represent the occurrence and effects of faults in system components, and specific models to precisely capture failure definitions and availability requirements. Based thereon, we show how the system's conformance to availability requirements can be verified automatically using the probabilistic model-checker Prism. To facilitate applicability of the approach, we provide basic building blocks for common fault occurrences, effects, and availability metrics. Moreover, we illustrate its applicability and benefits by evaluating the effects of using different positioning sensors on the availability of the Pick-and-Place Unit, a lab-scale demonstrator considered suitable for evaluations of novel research approaches. We argue that our approach is able to verify quantitative availability requirements while reducing redundant models, and thereby, potential inconsistencies.
AB - Availability is among the most important characteristics of manufacturing systems since it affects the productivity of the system. Yet, we argue that current approaches are inadequate for thoroughly analyzing availability of such systems. In this paper, we present a novel approach for analyzing availability of automated production systems. This approach extends the system's normative specification with deviation models to represent the occurrence and effects of faults in system components, and specific models to precisely capture failure definitions and availability requirements. Based thereon, we show how the system's conformance to availability requirements can be verified automatically using the probabilistic model-checker Prism. To facilitate applicability of the approach, we provide basic building blocks for common fault occurrences, effects, and availability metrics. Moreover, we illustrate its applicability and benefits by evaluating the effects of using different positioning sensors on the availability of the Pick-and-Place Unit, a lab-scale demonstrator considered suitable for evaluations of novel research approaches. We argue that our approach is able to verify quantitative availability requirements while reducing redundant models, and thereby, potential inconsistencies.
KW - Industrial Automation
KW - Model-Based Systems Engineering
KW - Probabilistic Model Checking
UR - http://www.scopus.com/inward/record.url?scp=85036450795&partnerID=8YFLogxK
U2 - 10.1145/3127041.3127051
DO - 10.1145/3127041.3127051
M3 - Conference contribution
AN - SCOPUS:85036450795
T3 - MEMOCODE 2017 - 15th ACM-IEEE International Conference on Formal Methods and Models for System Design
SP - 46
EP - 55
BT - MEMOCODE 2017 - 15th ACM-IEEE International Conference on Formal Methods and Models for System Design
PB - Association for Computing Machinery, Inc
Y2 - 29 September 2017 through 2 October 2017
ER -