Abstract:Industrie 4.0 environments generate an unprecedented amount of production data. This is due to the rising number of sensors and interconnected devices capable of emitting data in millisecond frequencies. Streaming analytics offers promising methodologies that can support handling and analysis of data volume and variety. Transparency and control over real-time data can increase production efficiency in tightly connected machine environments. Data transparency may avoid time-consuming assessment of machines to d… Show more
“…Context-specificity results from mechanisms used to analyze and evaluate production data regarding adherence to expectation. In related studies, three categories for design principles are identified regarding real-time anomaly detection evaluation, namely timeliness, threshold setting and qualitative anomaly assessment as presented in the following [7,13,18].…”
Section: Evaluation Categories For Real-time Anomaly Detection In Industrie 40mentioning
confidence: 99%
“…According to [13], timeliness is of major importance as optimal algorithms are supposed to detect anomalies as early as possible, so that countermeasures against anomalies can be initiated as soon as possible. The structured literature review presented in [18] identifies requirements for real-time anomaly detection in Industrie 4.0. The majority of analytical requirements such as fast data preparation emphasizes the importance of timeliness.…”
Section: Evaluation Categories For Real-time Anomaly Detection In Industrie 40mentioning
confidence: 99%
“…Among these, selfdiagnosis and self-repair include the autonomous detection and elimination of anomalies during production processes. Since about 2013, the number of publications on automated real-time anomaly detection has been increasing continuously [18]. The year 2021 ushers in the second decade of the Industrie 4.0 vision and self-diagnosis and self-repair competencies continue to form essential components [2].…”
Section: Introductionmentioning
confidence: 99%
“…There are several domain independent openly accessible algorithms that aim to detect anomalies in such streams [16]. Usually, several algorithms are deployed in parallel to determine anomalies as there is no single algorithm that fits all scenarios best [18]. To the best of our knowledge no Industrie 4.0 specific factors have been defined to support the selection of fitting real-time anomaly detection algorithms.…”
Section: Introductionmentioning
confidence: 99%
“…To find design principles, we conducted qualitative interviews with industry experts to get an understanding on what matters to them regarding anomaly detection evaluation. We restricted the scope of investigation to the three categories timeliness, threshold setting and qualitative classification as these are discussed in related studies [7,13,18].…”
The vision of Industrie 4.0 includes the automated reduction of anomalies in flexibly combined production machine groups up to a zero-failure ideal. Algorithmic real-time detection of production anomalies may build the basis for machine self-diagnosis and self-repair during production. Several real-time anomaly detection algorithms appeared in recent years. However, different algorithms applied to the same data may result in contradictory detections. Thus, real-time anomaly detection in Industrie 4.0 machine groups may require a benchmark ranking for algorithms to increase detection results' reliability. This paper makes a qualitative research contribution based on ten expert interviews to find design principles for such a benchmark ranking. The experts were interviewed on three categories, namely timeliness, thresholds and qualitative classification. The study's results can be used as groundwork for a prototypical implementation of a benchmark.
“…Context-specificity results from mechanisms used to analyze and evaluate production data regarding adherence to expectation. In related studies, three categories for design principles are identified regarding real-time anomaly detection evaluation, namely timeliness, threshold setting and qualitative anomaly assessment as presented in the following [7,13,18].…”
Section: Evaluation Categories For Real-time Anomaly Detection In Industrie 40mentioning
confidence: 99%
“…According to [13], timeliness is of major importance as optimal algorithms are supposed to detect anomalies as early as possible, so that countermeasures against anomalies can be initiated as soon as possible. The structured literature review presented in [18] identifies requirements for real-time anomaly detection in Industrie 4.0. The majority of analytical requirements such as fast data preparation emphasizes the importance of timeliness.…”
Section: Evaluation Categories For Real-time Anomaly Detection In Industrie 40mentioning
confidence: 99%
“…Among these, selfdiagnosis and self-repair include the autonomous detection and elimination of anomalies during production processes. Since about 2013, the number of publications on automated real-time anomaly detection has been increasing continuously [18]. The year 2021 ushers in the second decade of the Industrie 4.0 vision and self-diagnosis and self-repair competencies continue to form essential components [2].…”
Section: Introductionmentioning
confidence: 99%
“…There are several domain independent openly accessible algorithms that aim to detect anomalies in such streams [16]. Usually, several algorithms are deployed in parallel to determine anomalies as there is no single algorithm that fits all scenarios best [18]. To the best of our knowledge no Industrie 4.0 specific factors have been defined to support the selection of fitting real-time anomaly detection algorithms.…”
Section: Introductionmentioning
confidence: 99%
“…To find design principles, we conducted qualitative interviews with industry experts to get an understanding on what matters to them regarding anomaly detection evaluation. We restricted the scope of investigation to the three categories timeliness, threshold setting and qualitative classification as these are discussed in related studies [7,13,18].…”
The vision of Industrie 4.0 includes the automated reduction of anomalies in flexibly combined production machine groups up to a zero-failure ideal. Algorithmic real-time detection of production anomalies may build the basis for machine self-diagnosis and self-repair during production. Several real-time anomaly detection algorithms appeared in recent years. However, different algorithms applied to the same data may result in contradictory detections. Thus, real-time anomaly detection in Industrie 4.0 machine groups may require a benchmark ranking for algorithms to increase detection results' reliability. This paper makes a qualitative research contribution based on ten expert interviews to find design principles for such a benchmark ranking. The experts were interviewed on three categories, namely timeliness, thresholds and qualitative classification. The study's results can be used as groundwork for a prototypical implementation of a benchmark.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.