The MANTIS book : cyber physical system based proactive collaborative maintenance

Front Cover
Half Title Page
RIVER PUBLISHERS SERIES IN AUTOMATION, CONTROL AND ROBOTICS
Title Page
Copyright Page
Dedication Page
Contents
Acknowledgments
Foreword
List of Contributors
List of Figures
List of Tables
List of Abbreviations
Chapter 1 - Introduction
	1.1 Maintenance Today
	1.2 The Path to Proactive Maintenance
	1.3 Why to Read this Book
	References
Chapter 2 - Business Drivers of a Collaborative, Proactive Maintenance Solution
	2.1 Introduction
		2.1.1 CBM-based PM in Industry
		2.1.2 CBM-based PM in Service Business
		2.1.3 Life Cycle Cost and Overall Equipment Effectiveness
		2.1.4 Integrating IoT with Old Equipment
		2.1.5 CBM Strategy as a Maintenance Business Driver
	2.2 Optimization of Maintenance Costs
	2.3 Business Drivers for Collaborative Proactive Maintenance
		2.3.1 Maintenance Optimisation Models
		2.3.2 Objectives and Scope
		2.3.3 Maintenance Standards
		2.3.4 Maintenance-related Operational Planning
	2.4 Economic View of CBM-based PM
	2.5 Risks in CBM Plan Implementation
		2.5.1 Technology
		2.5.2 People
		2.5.3 Processes
		2.5.4 Organizational Culture
	References
Chapter 3 - The MANTIS Reference Architecture
	3.1 Introduction
		3.1.1 MANTIS Platform Architecture Overview
	3.2 The MANTIS Reference Architecture
		3.2.1 Related Work and Technologies
			3.2.1.1 Reference architecture for the industrial internet of things
			3.2.1.2 Data processing in Lambda
			3.2.1.3 Maintenance based on MIMOSA
		3.2.2 Architecture Model and Components
			3.2.2.1 Edge tier
			3.2.2.2 Platform tier
			3.2.2.3 Enterprise tier
			3.2.2.4 Multi stakeholder interactions
	3.3 Data Management
		3.3.1 Data Quality Considerations
		3.3.2 Utilization of Cloud Technologies
		3.3.3 Data Storages in MANTIS
		3.3.4 Storage Types
			3.3.4.1 Big data file systems
			3.3.4.2 NoSQL databases
	3.4 Interoperability and Runtime System Properties
		3.4.1 Interoperability Reference Model
		3.4.2 MANTIS Interoperability Guidelines
			3.4.2.1 Conceptual and application integration
			3.4.2.2 System interaction model
				3.4.2.2.1 MANTIS event model
				3.4.2.2.2 Patterns for interactions
			3.4.2.3 Implementation integration
	3.5 Information Security Model
		3.5.1 Digital Identity
		3.5.2 Information Mo
		3.5.3 Control Access Policy Specification
		3.5.4 Additional Requirements
	3.6 Architecture Evaluation
		3.6.1 Architecture Evaluation Goals, Benefits and Activities
		3.6.2 Concepts and Definitions
		3.6.3 Architecture Evaluation Types
	3.7 Conclusions
	References
Chapter 4 - Monitoring of Critical Assets
	4.1 The Industrial Environment
		4.1.1 Extreme High/Low Temperatures (Ovens, Turbines, Refrigeration Chambers etc.)
		4.1.2 High Pressure Environments (Pneumatic/Hydraulic Systems, Oil Conductions, Tires etc.)
		4.1.3 Nuclear Radiation (Reactors or Close and Near-By Areas)
		4.1.4 Abrasive or Poisonous Environments
		4.1.5 Presence of Explosive Substances or Gases
		4.1.6 Rotating or Moving Parts
	4.2 Industrial Sensor Characteristics
		4.2.1 Passive Wireless Sensors
		4.2.2 Low-Cost Sensor Solution Research
		4.2.3 Soft Sensor Computational Trust
	4.3 Bandwidth Optimization for Maintenance
		4.3.1 Reduced Data Amount and Key Process Indicators (KPI)
		4.3.2 Advanced Modulation Schemes
		4.3.3 EM Wave Polarization Diversity
	4.4 Wireless Communication in Challenging Environments
		4.4.1 Design Methodology Basis
		4.4.2 Requirement and Challenge Identification
		4.4.3 Channel Measurement
		4.4.4 Interference Detection and Characterization
		4.4.5 PHY Design/Selection and Implementation
			4.4.5.1 Single/multi carrier
			4.4.5.2 High performance/low power
		4.4.6 MAC Design/Selection and Implementation
			4.4.6.1 Real-time/deterministic MACs
			4.4.6.2 Low-power MACs
			4.4.6.3 High level protocols for error mitigation
		4.4.7 System Validation
			4.4.7.1 Channel emulation
			4.4.7.2 Performance tests
	4.5 Intelligent Functions in the Sensors and Edge Servers
		4.5.1 Intelligent Function: Self-Calibration
			4.5.1.1 Practical application: Press machine torque sensor
			4.5.1.2 Practical application: X-ray tube cathode filament monitoring
			4.5.1.3 Practical application: Compressed air system
		4.5.2 Intelligent Function: Self-Testing (Self-Validating)
			4.5.2.1 Practical application: Oil tank system
			4.5.2.2 Practical application: Air and water flow and temperature sensor
			4.5.2.3 Practical application: Sensors for the photovoltaic plants
		4.5.3 Intelligent Function: Self-Diagnostics
			4.5.3.1 Practical application: Environmental parameters
			4.5.3.2 Practical application: Intelligent process performance indicator
		4.5.4 Smart Function: Formatting
			4.5.4.1 Practical applications: Compressed air system
		4.5.5 Smart Function: Enhancement
			4.5.5.1 Practical application: Air and water flow and temperature
			4.5.5.2 Practical application: Railway strain sensor
			4.5.5.3 Practical application: Conventional energy production
		4.5.6 Smart Function: Transformation
			4.5.6.1 Practical application: Pressure drop estimation
		4.5.7 Smart Function: Fusion
			4.5.7.1 Practical application: Off-road and special purpose vehicle
			4.5.7.2 Practical application: MR magnet monitoring (e-Alert sensor)
			4.5.7.3 Practical application: MR critical components
	References
Chapter 5 - Providing Proactiveness: Data Analysis Techniques Portfolios
	5.1 Introduction
	5.2 Root Cause Failure Analysis
		5.2.1 Theoretical Background
		5.2.2 Techniques Catalogue
			5.2.2.1 Support vector machine
			5.2.2.2 Limit and trend checking
			5.2.2.3 Partial least squares regression
			5.2.2.4 Bayesian network
			5.2.2.5 Artificial neural network
			5.2.2.6 K-means clustering
			5.2.2.7 Attribute oriented induction
			5.2.2.8 Hidden Markov model
	5.3 Remaining Useful Life Identification of Wearing Components
		5.3.1 Theoretical Background
		5.3.2 Techniques Catalogue
		5.3.3 Physical Modelling
			5.3.3.1 Industrial automation
			5.3.3.2 Fleet’s maintenance
			5.3.3.3 Eolic systems
			5.3.3.4 Medical systems
		5.3.4 Artificial Neural Networks
			5.3.4.1 Deep neural networks
		5.3.5 Life Expectancy Models
			5.3.5.1 Time series analysis with attribute oriented induction
			5.3.5.2 Application to a pump
			5.3.5.3 Application to industrial forklifts
			5.3.5.4 Application to a gearbox
		5.3.6 Expert Systems
	5.4 Alerting and Prediction of Failures
		5.4.1 Theoretical Background
		5.4.2 Techniques Catalogue
			5.4.2.1 Nearest neighbour cold-deck imputation
			5.4.2.2 Support vector machine
			5.4.2.3 Linear discriminant analysis
			5.4.2.4 Pattern mining
			5.4.2.5 Temporal pattern mining
			5.4.2.6 Principal component analysis
			5.4.2.7 Hidden Semi-Markov model with Bayes classification
			5.4.2.8 Autoencoders
			5.4.2.9 Convolutional neural network with Gramian angular fields
			5.4.2.10 Recurrent neural network with long-short-term memory
			5.4.2.11 Change detection algorithm
			5.4.2.12 Fisher’s exact test
			5.4.2.13 Bonferroni correction
			5.4.2.14 Hypothesis testing using univariate parametric statistics
			5.4.2.15 Hypothesis testing using univariate non-parametricstatistics
			5.4.2.16 Mean, thresholds, normality tests
	5.5 Examples
		5.5.1 Usage Patterns/k-means
			5.5.1.1 Data analysis
			5.5.1.2 Results
				5.5.1.2.1 Plotting
				5.5.1.2.2 Replicability of results
				5.5.1.2.3 Summary of results
		5.5.2 Message Log Prediction Using LSTM
			5.5.2.1 Data interpretation and representation
				5.5.2.1.1 Litronic dataset
				5.5.2.1.2 Data representation
			5.5.2.2 Predictive models
			5.5.2.3 Results
				5.5.2.3.1 Evaluation of predictive models on small number of samples
				5.5.2.3.2 Evaluation of the ID-LSTM on OHE codes for more significant number of samples
			5.5.2.4 Discussion
		5.5.3 Metal-defect Classification
			5.5.3.1 Data collection
			5.5.3.2 Experiments
			5.5.3.3 Discussion
	References
Chapter 6 - From KPI Dashboards to Advanced Visualization
	6.1 HMI Functional Specifications and Interaction Model
		6.1.1 HMI Design Principle Followed in the MANTIS Project
		6.1.2 MANTIS HMI Specifications
			6.1.2.1 Functional specifications
			6.1.2.2 General requirements
		6.1.3 MANTIS HMI Model
			6.1.3.1 Functionalities supporting high level tasks
		6.1.4 HMI Design Recommendations
		6.1.5 MANTIS Platform Interface Requirements
			6.1.5.1 Analysis of different interface types
			6.1.5.2 PC HMI
		6.1.6 Recommendations for Platform Selection
			6.1.6.1 Web-based HMI
			6.1.6.2 Responsive design
		6.1.7 Interface Design Recommendations for MANTIS Platform
	6.2 Adaptive Interfaces
		6.2.1 Context-awareness Approach
			6.2.1.1 Context and context awareness fundamentals
			6.2.1.2 Context lifecycle in context-aware applications
			6.2.1.3 Adaptive and intelligent HMIs
			6.2.1.4 Context awareness for fault prediction and maintenance optimisation
			6.2.1.5 Context awareness for maintenance personalisation and decision-making
			6.2.1.6 Context awareness approaches in a proactive collaborative maintenance platform
		6.2.2 Interaction Based/Driven Approach
			6.2.2.1 Introduction
			6.2.2.2 Navigation tracking and storage
			6.2.2.3 Action logs
	6.3 Advanced Data Visualizations for HMIs
		6.3.1 Visualization of Raw Data
			6.3.1.1 Visualisation tools overview
			6.3.1.2 Scenario 1: Kibana
			6.3.1.3 Scenario 2: Textual and graphical data representation
		6.3.2 Augmented and Virtual Reality
			6.3.2.1 Scenario 1: Automated vibration monitoring
			6.3.2.2 Scenario 2: Condition and incoming maintenance alert for plant operators
	6.4 Usability Testing Methodology for Industrial HMIs
		6.4.1 Human-system Interaction – Usability Standards
			ISO 9241-11: Guidance on Usability (1998)
			ISO 9241-210: Ergonomics of Human-system Interaction – Part 210: Human-centred Design for Interactive Systems (2010)
			ISO/IEC 9126: Software Product Evaluation – Quality Characteristics and Guidelines for their Use (1991)
			ISO/IEC FDIS 9126-1: Software Engineering – Product Quality – Part 1: Quality Model (2000)
			ISO/IEC FDIS 9126-2: Software Engineering – Product Quality – Part 2: External Metrics and ISO/IEC FDIS 9126-3: Software Engineering – Product Quality – Part 3: Internal Metrics
			SO/IEC FDIS 9126-4: Software Engineering – Product Quality – Part 4: Quality in Use Metrics
			SO/TR 16982: Ergonomics of Human-system Interaction – Usability Methods Supporting Human-centred Design
		6.4.2 Usability Testing Methodology for MANTIS
	References
Chapter 7 - Success Stories on Real Pilots
	7.1 Shaver Production Plant
		7.1.1 Introduction to the Shaver Manufacturing Plant
		7.1.2 Scope and Logic
		7.1.3 Data Platform and Sensors
		7.1.4 Data Analytics and Maintenance Optimization
			7.1.4.1 Physical models and background
			7.1.4.2 Process monitoring with Principal Component Analysis & Hotelling’s T2
			7.1.4.3 Product quality prediction with partial least squares regression
			7.1.4.4 Computational trust
		7.1.5 Visualization and HMI
		7.1.6 Maintenance and Inventory Optimization Results
		7.1.7 Conclusions
	7.2 Deploying an User Friendly Monitoring System for Pultrusion Line Production
		7.2.1 Introduction to the Pultrusion Use Case
		7.2.2 Scope and Logic
		7.2.3 Data Platform and Sensors
		7.2.4 Human Machine Interfaces
		7.2.5 Maintenance Optimization and Validation Results
			7.2.5.1 Temperature control system located in the mixing area and in the storage area
			7.2.5.2 Cooling system for the injection chamber
			7.2.5.3 Compressed air system from pulling system
	7.3 Maintenance in Press Forming Machinery
		7.3.1 Introduction
		7.3.2 Scope and Logic
			7.3.2.1 Background information on the press machine
			7.3.2.2 Background information on the clutch brake component
		7.3.3 MANTIS Solutions for Press Machine
			7.3.3.1 Maintenance cloud platform
				7.3.3.1.1 Solution approach
				7.3.3.1.2 Results
			7.3.3.2 Torque measurement using wireless sensors
				7.3.3.2.1 Solution approach
				7.3.3.2.2 Results
			7.3.3.3 Head structural health monitoring
				7.3.3.3.1 Solution approach
				7.3.3.3.2 Results
			7.3.3.4 Bushing status measurement
				7.3.3.4.1 Solution approach
				7.3.3.4.2 Results
			7.3.3.5 Gears wear measurement
				7.3.3.5.1 Solution approach
				7.3.3.5.2 Results
			7.3.3.6 Press forces measurement
				7.3.3.6.1 Solution approach
				7.3.3.6.2 Results
		7.3.4 MANTIS Solutions for Clutch Brake
			7.3.4.1 Maintenance cloud platform by MGEP
				7.3.4.1.1 Background
				7.3.4.1.2 Solution approach
				7.3.4.1.3 Results
			7.3.4.2 Maintenance cloud platform by Tekniker
				7.3.4.2.1 Background
				7.3.4.2.2 Solution approach
			7.3.4.3 Friction material slippage
				7.3.4.3.1 Solution approach
				7.3.4.3.2 Results
			7.3.4.4 Brake spring degradation
				7.3.4.4.1 Solution approach
				7.3.4.4.2 Results
			7.3.4.5 Friction material wear
				7.3.4.5.1 Solution approach
				7.3.4.5.2 Results
			7.3.4.6 Piston chamber air leakage
				7.3.4.6.1 Solution approach
				7.3.4.6.2 Results
	7.4 Fault Detection for Metal Benders
		7.4.1 Introduction to Press Braking
		7.4.2 Design & Implementation
			7.4.2.1 Data collected by the machine’s sensors
			7.4.2.2 Wired nodes: The oil sensor
			7.4.2.3 Wireless nodes: The accelerometer
			7.4.2.4 Edge gateway
			7.4.2.5 Communication in the cloud
			7.4.2.6 Components for data analysis
			7.4.2.7 Human machine interface
		7.4.3 Data Analysis
			7.4.3.1 Data pre-processing
			7.4.3.2 Failure detection
				7.4.3.2.1 Parametric models
				7.4.3.2.2 Non-parametric models
				7.4.3.2.3 Evaluation and interpretation
		7.4.4 Conclusions
	7.5 Off-road and Special Purpose Vehicles
		7.5.1 Introduction to the Use Case on Vehicles
		7.5.2 Scope and Logic
		7.5.3 Data Platform and Sensors
		7.5.4 Data Analytics and Maintenance Optimization
		7.5.5 Conclusions
	7.6 Proactive Maintenance of Railway Switches
		7.6.1 Introduction to Railway Monitoring
		7.6.2 Scope and Logic
		7.6.3 Data Processing
		7.6.4 Measurement System for Proactive Maintenance of Railway Switches
			7.6.4.1 New factors collected
				7.6.4.1.1 Platform level
		7.6.5 Data Visualization
		7.6.6 Conclusion
	7.7 Fault Detection for Photovoltaic Plants
		7.7.1 Introduction to PV Plants
		7.7.2 Practical Application of Root Cause Analysis in Photovoltaic Plants
	7.8 Conventional Energy Production
		7.8.1 Introduction to the Plant Under Study
		7.8.2 Scope and Logic
		7.8.3 Monitoring Rolling Element Bearings
		7.8.4 IoT-Ticket Platform
		7.8.5 nmas Measuring System
		7.8.6 Mantis Cloud Platform
		7.8.7 Data Analytics and Maintenance Optimization
		7.8.8 Conclusions
	7.9 Health Equipment Maintenance
		7.9.1 Introduction to Health Imaging Systems
			7.9.1.1 Introduction to magnetic resonance
			7.9.1.2 Introduction to IGT systems
		7.9.2 Data Platform
		7.9.3 Data Lake
		7.9.4 ETL Scripts
		7.9.5 Data Warehouse
		7.9.6 Sensors
		7.9.7 Analysis and Decision Making Functionalities
			7.9.7.1 Predictive model deployment and live scoring
			7.9.7.2 Log pattern finder data
			7.9.7.3 Data sources
			7.9.7.4 Inspect and normalize the data
			7.9.7.5 Data pre-processing
			7.9.7.6 Data representation
			7.9.7.7 Equivalent log patterns
			7.9.7.8 Log pattern selection problem
			7.9.7.9 Design decisions
			7.9.7.10 Output
			7.9.7.11 Failure prediction
			7.9.7.12 Physical modeling
			7.9.7.13 Maintenance and inventory optimization
			7.9.7.14 Model and analysis
			7.9.7.15 Results and insights
			7.9.7.16 Visualization and HMI
			7.9.7.17 E-Alert portal
			7.9.7.18 Remote monitoring dashboard
			7.9.7.19 Conclusions
	References
Chapter 8 - Business Models: Proactive Monitoring and Maintenance
	8.1 Maintenance Present and Future Trends
		8.1.1 Tools
			8.1.1.1 Total productive maintenance
			8.1.1.2 Root-cause analysis
			8.1.1.3 Reliability centered maintenance
			8.1.1.4 Improving operational reliability
			8.1.1.5 Criticality analysis
			8.1.1.6 Risk-based maintenance
			8.1.1.7 Maintenance optimization models
			8.1.1.8 Model-based condition monitoring
		8.1.2 Trends
			8.1.2.1 Servitization
			8.1.2.2 Degree of automation
			8.1.2.3 Top-down vs. bottom-up
			8.1.2.4 Smart products
			8.1.2.5 Machine learning
	8.2 Shift to a Proactive Maintenance Business Landscape
		8.2.1 Key Success Factors
			8.2.1.1 User experience
			8.2.1.2 Privacy
			8.2.1.3 Scalability
			8.2.1.4 Technical debtThe surrounding infrastructure
			8.2.1.5 Skills
			8.2.1.6 Organizational structure
			8.2.1.7 Technology
	8.3 Proactive Maintenance Business Model
		8.3.1 Competitive Advantage for Asset Manufacturers
		8.3.2 Competitive Advantage for Asset Service Providers
		8.3.3 Competitive Advantage for Asset End Users
		8.3.4 Value Chains
		8.3.5 Main Technological and Non-technological Barriers/Obstacles for the Implementation
			8.3.5.1 Technological barriers
			8.3.5.2 Non-technological barriers
	8.4 From Business Model to Financial Projections
	8.5 Economic Tool to Evaluate Current and Future PMM Business Model
		8.5.1 Incomes Items
		8.5.2 Cost Items
		8.5.3 Schema of Economic Evaluation and Projection Report
	8.6 Railways Use-Case Financial Business Model
		8.6.1 Financial Business Benefits Within a Specific Railway Maintenance Solution
	8.7 Conclusions
	References
Chapter 9 - The Future of Maintenance
	9.1 Is it Cybernetic or Is it Human?
	9.2 Real-time Communication in Maintenance?
	9.3 How to Determine Granularity in Space and Time?
	9.4 Open or Closed Maintainability?
	9.5 Insourcing or Outsourcing?
	9.6 Explicit Modeling or Data-driven Pragmatics?
	9.7 How to Apply Virtual Reality and Augmented Reality?
	9.8 Service Robotics for Maintenance?
	9.9 How will the Maintenance Practices Change
	9.10 Conclusion
	References
Index
About the Editors
Back Cover