Search

The growing need for autonomous systems in offshore industries has contributed to the increased use of machine learning methods. These systems promise to improve safety in operations. However, the methods as enablers of autonomy are susceptible to various failures while interpreting data and making decisions. Several studies have highlighted the lack of research on the reliability and resilience of autonomous systems powered by these standard methods. Recent research provides sets of data interpretation methods. Despite the popularity of machine learning, there is a significant drop in knowledge when these methods result in failures. These failures further support autonomous systems in making wrong decisions. For autonomous systems, resilience and safety management should be an integrated functionality for recovery from risky situations and reporting of incidents. This research proposes an overview of machine learning methods for interpreting sensor data captured by drones operated manually and autonomously. We apply Isolation Forest for anomaly detection analysis and evaluate the Decision tree, Random forest, kNN, Logistic Regression, SVM, and, Naive Bayes for classification analysis. The methods are chosen based on their adequacy and comparative research prevalence. Comparison between the two drone operation modes contributes to understanding the reliability level for autonomously collected data. This research’s results provide an evaluation of machine learning methods’ performance across sensor data.

A pressure relief valve (PSV) is a key safety barrier to prevent the catastrophic rupture of pressure equipment in a process plant. The safety function of a PSV is to open and relieve the pressure when the equipment pressure exceeds the predefined set point. To achieve the desired availability of the PSV function, periodic function testing is regularly performed to confirm the correct functioning of a PSV. If a fault of the PSV function is detected by a function test, the PSV is repaired to a functioning state. For this reason, the interval between function tests has a direct influence on the probability of failure on demand (PFD) of the PSV function. On the other hand, unwanted leakage can occur due to human errors made during the preparation prior to a test and the reinstatement after the test. Such leakage is not desired due to the potential for being ignited and causing a major accident, but this aspect is often not considered in the availability assessment of PSVs. Therefore, this paper suggests a multi-phase Markov approach that can estimate the PFD of a PSV as well as the frequency of the leaks induced by the periodic tests. The suggested approach may be suitable for supporting the decision about the test interval for a PSV, considering both reliability and risk effect of extending the function test interval.

The Unmanned Autonomous Systems (UAS) are anticipated to have a permanent role in offshore operations, enhancing personnel, environmental, and asset safety. These systems can alert onshore operators of hazardous occurrences in the environment, in the form of anomalies in data, during real-time inspections, enabling early prevention of hazardous events. Time series data, collected by sensors that detect environmental phenomena, enables the observation of anomalous data as dynamic instances of the dataset. Recent research characterizes anomalies in terms of their patterns of occurrence in data. However, there is insufficient research on anomalous temporal change patterns. In this paper, we examine anomalies in relation to one another and propose a conceptual categorization system for anomalies based on their temporal changes. We demonstrate the categorization through a case study of potentially hazardous occurrences observed by UAS during underwater pipeline inspection. Analyzing anomalies based on their behavior can provide further information about current environmental changes and enable the early discovery of unwanted events, simultaneously minimizing false alarms that overwhelm the systems with low-significance information in real-time.

Formålet med denne rapporten er å gi næringen økt forståelse av pågående digitalisering, status og utfordringer, og hvordan denne utviklingen bør styres videre. Denne rapporten er en av seks SINTEF-rapporter fra prosjektet: "IKT-sikkerhet – Robusthet i petroleumssektoren 2020". Prosjektet har innhentet kunnskap om risiko, sårbarheter og IKT-sikkerhet for industrielle IKT-systemer. ISBN: 978-82-14-06481-0

Formålet med denne rapporten er å undersøke hvilke datakilder og data som benyttes i industrielle IKT-systemer og hvordan data behandles og prosesseres før de gjøres tilgjengelig i kontornettet. Styrker og sårbarheter knyttet til datakvalitet og sikring av data blir diskutert. Denne rapporten er en av seks SINTEF-rapporter fra prosjektet: "IKT-sikkerhet – Robusthet i petroleumssektoren 2020". Prosjektet har innhentet kunnskap om risiko, sårbarheter og IKT-sikkerhet for industrielle IKT-systemer. ISB: 978-82-14-06477-3