Pre-Conference Tutorial

Mr. DanielBalos is a Senior Consultant at Materials Testing Institute, University of Stuttgart, Germany and RBI/Integrity manager with TUV Rheinland Doha, specializing in Asset Integrity, Risk-Based inspection, Fitness for service and Damage mechanisms. MSc in Mechanical Engineering with the specialization in applicative IT and industrial management, PhD in application of data mining techniques on material behavior modelling for high temperature components.

More than 26 years of work in research and industrial projects, as well as training activities especially in risk- based inspections for power plants and gas and refining industry. Participated or led more than 20 EU funded projects and participated in a number of national projects in the area of material research and education abroad.

In these projects, a vast understanding and knowledge about materials, material degradation mechanisms, inspection methods, and risks has been accumulated. Project and risk management skills are proven in numerous projects in more than 20 years of project coordination and co-coordination.

In the area of Asset Integrity and RBI, he is active for more than 23 years, starting with participation in the key EU project in the area – RIMAP (Risk based inspection and maintenance procedures for European industry) – work in development and implementation of RBI approach for power plants, work in CEN CWA 15740 (standardization initiative for RBI in Europe), developed and Insert your latest photo implemented tools for RBI assessment of refining equipment in various projects. Lead in RBI and asset integrity projects in Kuwait and Qatar.

Participation and lead in the implementation project of RBI for NIS Serbia, EnBW Germany, as well as MOL Hungary, INA Croatia, ESKOM South Africa, Qatar Energy Qatar, Qatar Steel, SINOPEC China, KNPC Kuwait, OMV Petrom and Lukoil Romania and many others. Teaching RBI techniques and holding courses in RBI for petrochemical and power industry since 2005, with successful courses delivered in Germany, the Netherlands, Serbia, Romania, China, Vietnam, Myanmar, Egypt, Turkey, Tanzania and Malaysia.

Risk-based-inspection (RBI) has, in the last quarter of century, became the state-of-the-art approach, heavily utilized for inspection and maintenance planning and asset integrity management in oil and gas, refining and petrochemical industries. RBI represents the approach that links the operation-driven aspects– consequences of the failures with the asset integrity management strategies, one of the most important being inspection planning.

RBI methodology identifies the most critical static assets in the plant and helps focus the inspection efforts on them, allowing proper and balanced decision making towards inspection frequency, extent, and most efficient methods for damage monitoring in time.

This lecture aims to briefly present the methodology of the risk-based inspection approach as defined in the API RP 580/ASME PCC-3 and the RBI implementation methodology as defined in the API RP 581. The aim of the course is to guide the participants to understand the safety benefits of the RBI implementation and, on the other side, also appreciate the economic aspects of the risk. Calculation procedures for both probabilities and consequences will be discussed, and simple illustrative worked examples will be presented, with the aim to demonstrate the RBI process and dealing with risk and safety in the plant operation.

Dr. Durga Rao Karanki is currently working as a Senior RAMS Expert at Swiss Federal Railways (SBB CFF FFS) since Jan 2023. Earlier worked as a RAMS Manager at Siemens Mobility AG 06.2017- 12.2022, responsible for RAMS Management of European Train Control System projects. Prior to switching to railway industry, he had worked as a Scientist at Paul Scherrer Institute (ETH Board, Switzerland) from 2009 to 2017. His research at PSI primarily focused on dynamic safety assessment and uncertainty management. Prior to joining PSI, he worked as a Scientific Officer (2002- 2009) at Bhabha Atomic Research Centre (India), where he conducted research on dynamic fault tree analysis, uncertainty analysis, and risk informed decision making of nuclear power plants. He is also a visiting faculty at several technical institutes.

He has actively been involved in research and development of Reliability and Safety methods and their applications for the last 20 years. His work resulted in more than 70 publications including 4 books, 15 first author journal papers, and several conference papers, with more than 1500 citations. He received two awards for research Excellency from Society for Reliability Engineering, Quality and Operations Management (SREQOM). He is on the editorial board of three international journals in the area of reliability and risk analysis. He holds B. Tech (2000) in Electrical and Electronics Engineering from the Nagarjuna University (India), M. Tech (2002) in Reliability Engineering of Reliability Engineering Centre of Indian Institute of Technology (IIT) Kharagpur and Ph.D. from the IIT Bombay.

This presentation primarily focuses on RAMS-Management for Railway Systems. The current industrial practice, the standards, and the methods are briefly introduced. A few common issues in the practical applications of RAMS in the industry are brought up and their possible solutions are further discussed. During the life cycle of railway systems, there are several tasks related to Reliability, Availability, Maintainability, and Safety (RAMS), which ensures fulfillment of various requirements from national as well as European agencies besides specific customer requirements. Typical RAMS activities include apportionment of RAMS requirements, evaluation of design alternatives, hazard analysis, maintenance planning, operator support systems, RAM demonstration, safety case, FRACAS, etc. This presentation includes an overview of identified practical issues while executing railways projects and their potential solutions. For example, the issues in safety analysis include how to trade off requirements to be forwarded to customer/suppliers, defining test intervals, common cause factors while fulfilling tolerable hazard rates. Regarding reliability/availability analysis, estimation of model parameters in complex boundary conditions for hot standby systems and concerning maintainability issues, the demonstration tests and their optimization are challenging problems. All these issues and their possible solutions are illustrated with simple examples from the European Train Control System (ETCS), which is the signaling and control subsystem of rail traffic management system and give foundation for future automatic operations.

Key Words: RAMS, Hazard Analysis, Tolerable Hazard Rates, Safety Case, RAM Demonstration, Requirements Engineering, Railways, Signaling, European Train Control System.

Paluri S. V. Nataraj is a (Retd.) Professor of Systems and Control Engg Group at IIT Bombay. He joined the faculty of the Systems and Control Engineering Group at IIT Bombay in 1988. He was involved in teaching and research for more than 36 years at IIT Bombay. He was invited as a Visiting Professor to Spain in 2005, and as a visiting Professor to France in 2010 and 2015. His research interests were in the areas of Deep learning, Global Optimisation, Process Control, and Robust Control.

 Prof. Nataraj had conducted many courses for various industries and organisations, such as nuclear power, space and defence research organisations. He worked on several projects for these organisations and other industries. He gave numerous invited talks in North and South America, Europe, and Africa in areas such as optimisation, reliable computing, and control. Prof. Nataraj was the chairperson of several international conferences sponsored by IEEE and other societies. He also chaired and co-chaired several sessions in various international conferences in India and abroad.

This talkgives a brief introduction to ML and DL, followed by a description of two engineering applications of the AI based methods for modelling and fault diagnosis. The applications considered are DC motor and laboratory gas turbine engine.

Distinguished Staff Scientist – Idaho National Laboratory, Idaho Falls, ID 83404

B.E. University of Madras, Chennai, India, (2001); M.S. University of Tennessee, Knoxville (2005); and the Ph.D. Purdue University (2011). In present position, his current research advances sensor technologies, artificial intelligence (AI)/machine learning (ML), autonomous controls, and wireless communications in (1) achieving automation for efficiency gain in current fleet of operating reactors; (2) developing real-time predictive control to achieve autonomous control and operation of advanced reactors; (3) performing of risk-informed predictive evaluation; and (4) formulating an approach to address scalability, explainability, and trustworthiness of AI/ML technologies. He received the 2023 Asian-American Most Promising Engineer of the Year award for his notable technical contributions and community services. Other recognitions include 2015 Presidential Early Career Awards for Scientists and Engineers; 2016 Laboratory Director Early Career Achievement Award; and 2019 American Nuclear Society (ANS) Human Factors Instrumentation and Control Division (HFICD) Ted Quinn Early Career Award. Over 100 peer-reviewed publications, two patents, and six software copyrights. He is a Member of the ANS since 2011. He served as an executive member of HFICD – 2015-2018 and is the Publication Chair of HFICD – since 2022. He was the Technical Program Chair for Nuclear Plant Instrumentation, Control, & Human-Machine Interface Technologies (NPIC&HMIT) 2023.

Professor (Dr.) Debabrata Datta is retired in the year 2019 from Bhabha Atomic Research Centre, Mumbai as Senior Scientist and Head of the Radiological Physics & Advisory Division. He has contributed in Nuclear and Radiation safety in various units of Department of Atomic Energy. He has provided his meritorious service in BARC and other Department of Atomic Energy for a period of 34 years (1985-2019). At present, he is associated with Heritage Institute of Technology, Kolkata in the capacity of Joint Director Research & Development and Professor in the department of Information Technology. He is also an adjunct professor in the department of Mathematics of Royal Global University, Assam and SRM Institute of Science & Technology, Kattankulathur, Chennai. His academic excellence is: Master of Science (Nuclear Physics) from Calcutta University (1981-1983), and Master of Philosophy (High Energy Nuclear Physics) from Calcutta University (1983-1984), Graduate from OCES course of BARC Training School (1984-1985) and PhD (Science) from Mumbai University in 2000. He has worked in the field of Artificial Intelligence and developed an Expert System for Nuclear Industry. He has worked in diverse field of science and contributed in the field of Soft Computing, Mathematical and Statistical modeling, Uncertainty modeling, Risk analysis, Molecular Dynamics simulation. He has developed Lattice Boltzmann method to study the design of geological repository for disposal of high-level radioactive waste. He has developed many applications software, algorithms and designed radiation shielding of radiochemical plants. Some of his contributions are, (i) Design of Induction furnace for characterization of borosilicate glass for managing radioactive waste, (ii) Lattice Boltzmann method for fluid flow analysis through micro-porous materials, (iii) processing of medical image using soft computing, (iv) numerical simulation of insulin delivery system for Type-II diabetic patient, (v) optical communication using laser, (vi) molecular dynamics simulation of material processing using diamond turning machine, plasma cutting using electric discharge machining process and (vii) machine learning software to handle water quality data (big data analysis) of Karnataka and air quality of Delhi. He has also contributed in quality control of radiotherapy machines (CT, MRI, PET-SCAN) installed and used in various hospitals of our country. He has been deputed to International Atomic Energy Agency (IAEA), Vienna, Austria (2004) to test the quality control and diagnose of agency developed software RAIS (regulatory authority information system) and in Indonesia (2017) to strengthen the medical physics education by the e-learning system software AMPLE. He has also contributed in Defence sector as radiological advisor/regulatory inspector in the field of Industrial Radiography and in hospital as Medical Physicist. He has got 34 years of experience in software development as well as in R&D of diverse fields. His present research is focussed on AI/ML techniques in healthcare system, quantum machine learning and quantum computing. He has been recognized by Marquis WHO’s WHO in the world continuously from 2015-2020 and he is the recipient of national award ‘Millennium Plaques of Honour’, from Indian Science Congress Association (ISCA) in 2010, Group Achievement Award for Science & Technological Excellence from BARC in 2018, JALVIGYAN PURASKAR (BEST RESEARCH PAPER Award) – from ISH Journal of Hydraulic Engineering in 2018 and other meritorious awards from many reputed scientific societies. He has supervised/co-supervised 7 PhD thesis, 10 M. Tech thesis, examined 12 PhD Thesis and 30 M. Tech thesis. He has executed 12 BRNS projects as principal collaborator. He has got more than 250 peer reviewed international journal publications and “three patents (2020-2022)” to his credit and delivered a large number of invited lectures, keynote lectures in many International and national conferences. He is also an editor, associate editor and editorial board member of many international and national journals. He is life member of many international and national scientific bodies.

Remaining Useful life (RUL) is defined as the duration of the time period between the present and the end of a system’s useful life. Real time monitoring of damage /degradation of an engineering system necessitates the prognosis and health management (PHM). The state of health of any system under study during ageing of its various features can be predicted by RUL which is basically depends of conditional maintenance. An accurate estimate of RUL can appropriately prescribe the state of health of the system, which can be anything say either, an industry, or a specific instrument, or supply chain management system or the system can be healthcare system. Obviously, features of the system are time dependent and have a substantial interaction with the environment. Traditional data driven model of RUL fails very often not only due to complexity of the system but also due to misunderstandings of the behaviour of the system, especially time dependent interrelation of the features. Machine learning (ML) algorithm can be applied to formulate a suitable model for predicting RUL and that type of model is always better than data driven model because ML driven model is always dynamic and tome to time it always updates the features of the system. ML pipeline and ensemble learning algorithm such as Gaussian Naïve Bayes, Decision Tree, Random Forest, AdaBoost, Gradient Boost, Extreme Gradient Boost and deep learning-based CNN will be presented to design the computational model of RUL. This tutorial will focus the step-by-step guidance to design the predictive model of RUL using ML algorithms as mentioned. Two case studies, one from an engineering system and other from healthcare system will be demonstrated.

Keywords: RUL, ML model, Random forest, data driven, AdaBoost