...
When | Who | What |
---|---|---|
19.01.2018 | Behnaz Hosseinnia Xiaopeng Li | Title: New Ph.D. Candidate Self-introduction Speaker: Behnaz Hosseinnia Self intorduction_Behnaz.pdf Abstract: Introduce herself and talk about her previous research project at Safety and Security Science group at TU Delft, the Netherlands. She will also give a brief introductionofherPhD research topic at RAMS group. About Speaker: NewPhD student within RAMS group. She received her master degree in chemical engineering (minor: HSE) and did her undergraduate studies in Safety and Technical Protection Engineering in Petroleum University of Technology (Iran) with the first rank honor. In her M.Sc.thesis she has studied the application of fire and explosion modeling, failure probability analysis and quantitative risk-based decision making in petroleum pipelines emergency response planning. Besides the academic experience, two internships with National Iranian Oil Company (NIOC) and four-year work with engineering consultancies have enabled her to combine the technical knowledge with hands-on experience in risk analysis, process safety and accidents consequence modeling. From 2016, she worked as a researcher at the Safety and Security Science group in TU Delft, the Netherlands. Her research project was funded by LDE (Leiden. Delft. Erasmus) Centre for Safety and Security and focused on developing an effective decision-making tool for multi-plant emergency response planning against terrorist attacks in the chemical industrial areas. Speaker: Xiaopeng Li ( Exchange Ph.D. candidate) Self-introduction_Xiaopeng Li.pdf About Speaker: He is from College of Management and Economics, Tianjin University, China. He will introduce himself and his previous research works on warranty management in Tianjin University and his research plan in NTNU. |
26.01.2018 | Wenyan Song | Title: Data-driven fuzzy modeling method and its application Data driven_Wenyan.pdf Abstract: Fuzzy modeling method is an effective computation intelligence technology that can handle data information and human knowledge. Now it has been widely applied in system modeling, decision system design and data analysis. In this talk, we will introduce some basic concepts about fuzzy set theory and a few examples of fuzzy modeling method combined with other machine learning algorithms. Besides, we also introduce some application cases about fuzzy modeling. Speaker: Dr. Wenyan Song is an associated professor in School of Economic, Dongbei University of Finance and Economics University in China. She has research interests in machine-learning, and she works more in applying the technique to industrial process and financial market. |
02.02.2018 | HyungJu Kim | Title: Systems-Theoretic Process Analysis (STPA) for subsea systemsSTPA to Subsea_HyungJun Kim.pdf Abstract: Systems-Theoretic Process Analysis (STPA) is a recently developed hazard identification technique that is based on control and systems theory. Previous studies on STPA emphasizes two major strengths of the method: (1) STPA provides a systematic top-down approach that enables early identification of system flaws, and (2) STPA covers a wider scope of hazards compared to traditional methods. Despite these advantages, there are only a limited number of studies that have applied the method to subsea systems. It is therefore of interest to investigate how STPA can be used to formulate new or verify existing requirements for safety-critical systems for subsea facilities. The contents of the presentation are 1) Introduction to STPA; 2) Ongoing STPA studies in RAMS group; 3) STPA to subsea gas compression system - ESREL 2018 conference; 4) STPAtoisolationofsubsea wells - OTC 2018 conference About speaker: Hyungju Kim is a postdoctoral fellow at RAMS group, MTP, NTNU. His research topic is new control and safety philosophies of subsea systems, which is a part of SFISUBPRO project. He completed his Ph.D. in RAMS at NTNU and earned his Master's and Bachelor's degree in Naval Architecture and Ocean Engineering at Seoul National University. Before he started his Ph.D. at NTNU, he worked for Samsung Heavy Industries for seven years, as a naval architect. |
16.02.2018
| Harald Rødseth | Title: Risk-based maintenance backlog Maintenance backlog _Harald.pdf Abstract: TBAA relevant issue in manufacturing and production seems to be “silo”- organizations and “silo”-planning with lack of coordination between departments. Integrated Planning (IPL) is a concept that aims to cope with this “silo”-problem. With the ground-breaking potentials from Industry 4.0, it should be expected that the advancement of IPL will speed up in development and implementation in companies. To manage IPL sound key performance indicators (KPIs) must be implemented and established by the company. A promising indicator for IPL is maintenance backlog (MB). A strength of this indicator is the capability to be modeled with Risk OMT (Risk modeling – Integration of Organisational, human and Technical factors). It remains to investigate how MB can be modeled to a Quantitative Risk Analysis (QRA). The main objective of this article is to develop a model of MB in QRA. In particular, the article demonstrates a case study of a production system where both fault tree analysis (FTA), and event tree analysis (ETA) is modeled. The article discusses the demonstration results and evaluates how potentials in Industry 4.0 can support QRA. About speaker: Harald Rødseth, Postdoctoral fellow, MTP, NTNU. Opponents: Prof. Jørn Vatn & Ph.D. candidate Yun Zhang |
02.03.2017 | Prof. Anne Barros | Title: On the Use of Piecewise Deterministic Markov Processes in Reliability and Maintenance PDMP_Anne Barros.pdf Abstract: Piecewise deterministic Markov processes (PDMP) are widely used in dynamic reliability to model phenomena which are considered as deterministic most of the time (e.g. evolution of the fluid level in a tank) and which are influenced from time to time by stochastic events (e.g. failures in the control loop for the fluid level). Usually, a PDMP is made of a set of differential equations (deterministic part) whose solutions can experience random “jumps” (effect of stochastic events). We present here a very specific type of PDMP: the deterministic part is reduced to a set of trivial differential equations whose solutions will be used to measure the time elapsed since the last stochastic events. Hence, the deterministic part is not related to any physical phenomena but is an artifact to model stochastic behaviors that require a combination of discrete random “jumps” and continuous variables to count time. We will discuss how such PDMP can be used to study maintained systems with several units. Stochastic jumps will model failure, repair, detection times while the continuous variables will model deterministic repair durations or delays, the time between inspections, time spent in different states of interest (especially in case of Weibull lifetime or repair time). We will try to explore how such a formalism can help for the modeling work, and when an explicit numerical scheme can be easily developed to calculate the quantities of interest (MDT, Availability, Reliability, Mean number of repairs...). About speaker: Anne Barros is a professor at NTNU in the Department of Mechanical and Industrial Engineering. Her research activity is focused on the use of stochastic processes in Reliability and Maintenance. Career: Master's degree in Systems Control from the University of Technology of Compiegne UTC (France), 2000, Master of Engineering in Industrial System from the University of Technology of Troyes UTT (France), 2000andPhD(dr.philos) in Optimisation and System Safety from UTT, 2003. Associate Professor of Reliability and Maintenance engineering at UTT (2003-2011). Professor of Reliability and Maintenance engineering at UTT (2011-2014). Professor of Subsea Reliability at Norwegian University of Sciences and Technology NTNU (professorship founded by DNV-GL, 2014-). |
16.03.2018 | Prof. Antoine Rauzy | Title: Reliability Analysis of Looped Systems Abstract: It is well widely assumed (and true) that reliability block diagrams are equivalent to fault trees, although this equivalence is formally established in no textbook. A reliability block diagram can be seen as an oriented graph: blocks are represented by nodes and connections between blocks by edges of the graph. This graph is oriented, contains no loop (is acyclic), has a unique source node s and a unique target node t. The system described by the reliability block diagram works if and only if there is a working s-t path in the graph, an i.e. s-t path along which all nodes are working (only nodes can fail, edges are assumed to be perfectly reliable). An interesting question is: what does happen if we accept loops, i.e. if we consider reliability networks rather than reliability block diagrams? Assessing the reliability of networks is indeed of practical importance as most infrastructures can be seen as networks. Nevertheless, reliability networks are only seldom used as a modeling tool. In this seminar, the main mathematical results explaining why it is so and present several algorithms to solve the problem will be recalled. About speaker: Antoine B. Rauzy has currently a full professor position at Norwegian University of Science and Technology (NTNU, Trondheim, Norway). He is also the head of the chair Blériot-Fabre, sponsored by the group SAFRAN, at CentraleSupélec (Paris, France). During his career, he moved back and forth from academia to industry, being notably senior researcher at French National Center for Scientific Research (CNRS), associate professor at Universities of Bordeaux and Marseille, professor at Ecole Polytechnique and Ecole Centrale Paris, CEO of the start-up company ARBoost Technologies he founded, and director of the R&D department on Systems Engineering at Dassault Systemes (largest French software editor). He gothisPhDin1989 and his tenure (habilitation àdirigerdes Recherches) in 1996, both in computer science. He works in the reliability engineering field for more than 20 years. He extended his research topics to systems engineering more recently. He published over 150 articles in international conferences and journals. He is on the advisory boards of several international conferences and journals and is regularly invited to deliver keynote talks at international conferences. He renewed mathematical foundations and designed state-of-the-art algorithms for probabilistic safety/risk assessment. Alone or with his students and collaborators, he developed safety/risk assessment software that is daily used in industry (Aralia, XFTA, MarkXPR). He is also the main designer of the AltaRica modeling language and the scientific advisor of the Open-AltaRica project (IRT SystemX). He managed numerous collaborations between academia and industry, in Europe, in theUSAandinJapan, and has been the adviser of fifteenPhDtheses. |
06.04.2018 | Shenae Lee
| Title: Application of Bayesian Belief Networks (BBNS) for process plant Shenae RAMS seminar.pdf Abstract: This study is about risk analysis of process plants where main accidents continue to occur. Conventional risk analysis method has the limitation of having a static structure, while another challenge is difficult to aggregate operational data from different sources. To focus on these limitations, the paper suggests an approach based on Bayesian networks (BNs), and it is illustrated by a case study of a pressure relief valve in an Ammonia plant. The approach seems to be suitable for updating frequency of accident scenarios when new risk information is collected during the operational phase. About speaker: Shenae Lee is a Ph.D. student in RAMS group. The topic for herPhDis a dynamic risk analysis of major accident hazards in process facility operations to support safety-critical decisions. She is under the main supervision of Prof. Nicola Paltrinieri in RAMS group. She has a Master’s from the international RAMS program, NTNU. Opponents: Pierluigi Salvo Rossi, Kongsberg Digital AS, Norway; Juntao Zhang, RAMS group |
20.04.2018 | Assoc. Prof. Astrid S. de Wijn | Title: Criticality in Dynamic Arrest: Correspondence between Glasses and Traffic RAMS_Astrid.pdf Abstract: The dynamic arrest is a general phenomenon across a wide range of dynamic systems including glasses, traffic flow, and dynamics in cells, but the universality of dynamic arrest phenomena remains unclear. We connect the emergence of traffic jams in a simple traffic flow model directly to the dynamic slowing down in kinetically constrained models for glasses. Using the Nagel-Schreckenberg model to simulate traffic flow, we show that the emergence of jammed traffic acquires the signature of a sharp transition in the limit corresponding to overcautious driving. We identify a true dynamic critical point marking the onset of coexistence between free-flowing and jammed traffic and demonstrate its analogy to kinetically constrained glass models. We find diverging correlations analogous to those at a critical point of thermodynamic phase transitions. About speaker: Associate Professor at Department of MTP at NTNU and also attached to the department of Physics Stockholm University. Her interest focus on statistical mechanics, tribology, condensed matter, nonlinear dynamics, surface science. More detail please check: http://www.syonax.net/science/research.html |
04.05.2018 | Juntao Zhang
Jon Martin Fordal | Title: Adapting Systematic Theoretic Process Analysis for Reliability Analysis Juntao Zhang_RAMS seminar_04.05.pdf Abstract: The rather recent method named System-Theoretic Process Analysis (STPA) is one promising candidate to improve the coverage of hazard identification in complex complex systems that involve highly coupled parts, non-linear interactions, and software-intensive functionalities. Still, there is no guideline for utilizing STPA output to evaluate the potential of loss, which is important for a basis for decision-making about system configuration and equipment selection. The focus of this article is placing on the interface between STPA and reliability, availability and maintainability (RAM) modellingmodeling. The approach named STPA-RAM modelling modeling is proposed to translate feedback control loops into Petri-nets for discrete event simulation. The proposed approach is demonstrated with a simple case related to the subsea design concept. It has been found that the new proposed approach extends the application of STPA, while also improving, and as such reducing completeness uncertainty and model uncertainty, associated with input data and information for RAM modellingmodeling. About speaker: Juntao Zhang is a PhD candidate in RAMS group, in related to research center SUBPRO. The PhD topic is incorporating reliability and availability analysis in the early design phase of subsea systems. He is under the main supervision of Prof. Mary Ann Lundteigen in RAMS group. He has a Master degree from degree from RAMS program, NTNU. Opponents: Prof. Cecilia Haskins, PhD candidate Shenae Lee
Title: New Ph.D. Candidate Self-introduction Self_introduction_Fordal.pdf Abstract: Self-introduction of Jon Martin Fordal – background & and next steps About speaker: Jon Martin Fordal comes from Stjørdal and has a bachelor degree in mechanical engineering, master of science in industrial engineering, and has worked as a maintenance engineer in Elkem ASA. 1st of march he started as a PhD candidate, and will be working on the research project “CPS Plant”. The project will develop a framework for the Norwegian approach for the digital manufacturing industry. The consortium consists of 3 Norwegian industry partners, Norsk Hydro, Benteler Automotive and Hycast, and SINTEF Digital and NTNU (Trondheim and Gjovik) are the academic partners. SINTEF Raufoss Manufacturing is the project leader. |
11.05.2018 | Assoc. Prof. Viggo Gabriel Borg Pedersen | Title: Monitoring data quality – Demonstration of vibration measurements Abstract: The quality of condition monitoring data affects remaining useful life prognosis based on the monitoring data. The whole measurement chain has to be customized in such a way that relevant data of the right quality is collected for analysis purposes. Choosing the right sensor for the job and installing it correctly at the optimal position require knowledge of the process and unit being monitored. Conversion and transmission of sensor data is another source of error in data acquisition systems. The presentation will highlight and discuss some issues related to the above-mentioned challenges. Demonstration of gathering vibration data on two live test rig's is part of the presentation. The rigs of the type rotational equipment are mounted in the lecture room. About speaker: Assistant Professor teaching operation & maintenance and design of piping systems at bachelor level the last five years. 10 years’ work experience in construction, operation, and maintenance of offshore installations & ships. 9 of these as an engineer in different positions. 9 years’ work experience related to hydropower, operation, and maintenance, laboratory management in different engineering positions. MSc – Marine Technology, marine machinery; MA – Project management; BSc –mechanical engineering |
25.05.2018 | Yun Zhang | Title: Modelling methodology and patterns for the monitored system using formal modeling language AltaRica 3.0: a study case Abstract: TBA About speaker: Yun Zhang, Ph.D.student of RAMS group, MTP, NTNU. |
01.06.2018 | Himanshu Srivastav, Lin Xie, Aibo Zhang | Title: Presentation for ESREL Abstract: TBA About speaker: TBA |
08.06.2018 | Associate Professor Cecilia Haskins | Title: Systems Engineering in 45 minutes Abstract: TBA About speaker: TBA |
15.06.2018 | Closing social event |
...