Evolution Management of Production Facilities by Semi-Automated Requirement Verification

Abstract

The concept described in this contribution utilizes available process data in production systems to enable to verify the fulfillment of (non-functional) requirements during operation. Therefore, the contribution presents a systematic approach to automatically derive property values out of signal traces by using adaptable runtime models. Those can be checked for the violation of limit values in order to verify the fulfillment or violation of requirements on these properties during an evolution. In an example application to a Pick and Place unit, the concept is used together with an anomaly detection method to support the operator during the evolution process by constantly providing information regarding requirement fulfillment.

The CRI-model: A domain-independent taxonomy for non-conformance between observed and specified behaviour

“…Further, in contrast to anomaly detection, this contribution also explicit considers coevolution as deviations of the specified behaviour due to changes. Approaches like the one of Ladiges et al explicit consider such deviations detected in domain-specific models [13]. The taxonomy of changes used by Ladiges et al, further described in [14] and extended by Vogel-Heusser et al in [30,31], serves as a basis for categories of the CRI-Model regarding deviations regarding Cyber-Physical Systems in this contribution.…”

“…This is why in these long-living production system detection of deviations between the system and its specification are of high interest [21]. These detection approaches do not assume faulty behaviour directly, but rather initialise a decision process to decide if an unexpected deviation is an anomaly or a deviation due to old specifications [13]. Therefore, detection of anomalies and deviations have a strong overlap which is why the following section shows how the CRI-Model can be applied also for detection of deviations.…”

“…Due to its inherent interdisciplinary nature, detection of nonconforming behaviour is in cyber-physical production systems mostly model-based, which means that deviations are often identified by comparing the actual behaviour not just by metrics, but by specific behaviour models [13]. Here, domain-specific detection methods which observe the system behaviour on the software interface and comparing it with a model representation of the last known behaviour are used in order to support failure detection or evolution of the production systems [13]. To tackle the problem of deviations as well as deviation with anomalies, we divide evolution according to Vogel-Heusser et al in long-term planned evolution and short time ad-hoc evolution [30].…”

“…Detailed descriptions are available in [31] and in over 50 related publications (see e.g. [13,15,30]). The demonstrator includes 23 common evolution scenarios with 15 sequential and 8 parallel evolution steps which typically happen in production systems.…”

“…To observe deviations between the specification and observed behaviour this article uses a self-documentation approach that rest on generating signal-based behaviour models (see [13]). The approach automatically learns and analyses the behaviour based on externally observable signal events.…”