Computer Safety, Reliability, and Security. SAFECOMP 2022 Workshops: DECSoS, DepDevOps, SASSUR, SENSEI, USDAI, and WAISE Munich, Germany, September ... (Lecture Notes in Computer Science)

Preface
Organization
Contents
17th International ERCIM/EWICS/ARTEMIS Workshop on Dependable Smart Embedded Cyber-Physical Systems and Systems-of-Systems (DECSoS 2022)
En
	1 Introduction
	2 ECSEL and KDT JU: The European Cyber-Physical Systems Initiative
	3 This Year’s Workshop
	International Program Committee 2022
Triggering Conditions Analysis and Use Case for Validation of ADAS/ADS Functions
	1 Introduction
	2 Use Case and Methodology
		2.1 ADAS/ADS Function Description
		2.2 Operational Design Domain
		2.3 Triggering Conditions List
		2.4 Relevant Scenarios
		2.5 Hazard Analysis and Risk Assessment
		2.6 Qualitative Analysis
		2.7 Testing and Evaluation
		2.8 SOTIF Modifications
		2.9 Residual Risk Quantification
		2.10 SOTIF Argumentation
	3 Outlook
	References
A Statistical View on Automated Driving System Safety Architectures
	1 Functional Safety and Safety of the Intended Functionality
		1.1 Terms
	2 Tolerable Risk
	3 Current Sense-Plan-Act Architectures
	4 Challenge of Control and Protection System Architecture
		4.1 Probabilistic Target to Demonstrate
		4.2 Functional Safety Related to Systematic E/E-failures
		4.3 SOTIF - Low Demand Frequency Events
		4.4 SOTIF - High Demand Frequency Events
	5 Consequences
		5.1 Common Cause  Factor
		5.2 Architectural Consequence
		5.3 Quantitative Injury Risk Reduction Due to the Warning Subsystem
		5.4 Further Benefits of the Warning Subsystem
	6 Conclusion
	References
Building a Test Bed for the Periodical Technical Inspection of Sensors Perceiving the Environment in Autonomous Vehicles
	1 Introduction
	2 Periodic Technical Inspection in Germany
		2.1 Driver Assistance Systems in the Periodic Technical Inspection
	3 Sensor Verification in the Field
	4 Advancing the Periodical Technical Inspection
		4.1 Sensors for the Environmental Perception
		4.2 Test Bed for the Sensor Systems in the Periodical Technical Inspection
		4.3 Positioning of the Vehicle on the Test Bed
		4.4 Inspection of the Sensor Systems
		4.5 Positioning of Targets
		4.6 Sensor Targets for the Inspection
		4.7 Exemplary Test Performed in Regard to the Lidar System
		4.8 Required Information About the Vehicle Under Test
		4.9 Related Work and Differentiation
	5 Conclusion and Future Work
	References
Testing and Verification of the Deep Neural Networks Against Sparse Pixel Defects
	1 Introduction
	2 Related Work
	3 Proposed DES-based Approach
		3.1 Use of the DE
		3.2 Use of the DES
	4 Integration into Safety-Critical Systems Development
	5 Results
	6 Conclusions
	References
Observing the Impact of Multicore Execution Platform for TSP Systems Under Schedulability, Security and Safety Constraints
	1 Introduction
	2 Background and Assumptions
		2.1 System Model and Assumptions
		2.2 Security and Safety
		2.3 Multi-objective Optimization
	3 Design Space Exploration (DSE) Approach for Multicore TSP Systems
	4 Test Cases and Evaluation
	5 Related Work
	6 Conclusion
	References
Security Assessment Prospects as Part of Vehicle Regulations
	1 Introduction
	2 Regulatory Background
		2.1 Type Approval
		2.2 Periodic Technical Inspections (PTI)
	3 Cybersecurity Challenges over the Years of Vehicle Operation
	4 Related Work
	5 State of the Art – Security Standard and Regulations
		5.1 ISO/SAE 21434 – Cybersecurity Engineering
		5.2 UN R155 Cybersecurity Regulation
	6 Discussion and Future Assessment Prospects
	7 Conclusion
	References
UNECE Threat List Case Study: Prediction of Cyber Risks in the Automotive Domain Using Regression Analysis
	1 Introduction
	2 UNECE Automotive Potential Threats
		2.1 Estimation of the Likelihood Factors
		2.2 Estimation of the Impact Factors
		2.3 Building the Dataset
		2.4 Regression Analysis
	3 Case Study: UNECE Automotive Threats List
		3.1 Regression Approaches
		3.2 Results Evaluation
	4 Summary, Conclusion, and Future Work
	References
Watch: A Validation Framework and Language for Tool Qualification
	1 Introduction and Motivation
	2 Related Work
	3 The Watch Framework and Language
		3.1 End-To-End Testing of FuSa Relevant Toolchains
		3.2 White-Box Testing: Low-Level Testing of CI Jobs
	4 Conclusion and Outlook
	References
Criteria for the Analysis of Gaps and Limitations of V&V Methods for Safety- and Security-Critical Systems
	1 Introduction
	2 Background
		2.1 The VALU3S Project
		2.2 Related Work
	3 Criteria for the Analysis of Gaps and Limitations of V&V Methods
	4 Application of the Criteria for the Analysis of Gaps and Limitations of V&V Methods
		4.1 Application of the Criteria to Two Methods
		4.2 Application Results
		4.3 Discussion
	5 Conclusion
	References
3rd International Workshop on Dependable Development-Operation Continuum Methods for Dependable Cyber-Physical System (DepDevOps 2022)
En
	1 Introduction
	2 H2020 Projects: Dependable DevOps
	3 Acknowledgments
	4 International Program Committee
A Guided Search for Races Based on Data Flow Patterns
	1 Introduction
	2 Related Work
	3 Race-Prone Data Flow Patterns
		3.1 Terminology and Assumptions
		3.2 Pattern 1
		3.3 Pattern 2
		3.4 Pattern 3
	4 Pattern-Based Race Detection
		4.1 Phase 1: Pattern-Based Identification of Internal Races
		4.2 Phase 2: Analysis of Extendibility from Internal Races to System Races
	5 Evaluation
		5.1 Example
		5.2 Race Detection Capability and Effort
		5.3 Comparison with Random Search
	6 Conclusion
	References
9th International Workshop on Next Generation of System Assurance Approaches for Critical Systems (SASSUR 2022)
En
	1 Introduction
	2 This Year’s Workshop
	Workshop Committees
		Organization Committee
		Programme Committee and Reviewers
DevOps and Safety? SafeOps! Towards Ensuring Safety in Feature-Driven Development with Frequent Releases
	1 Introduction
	2 Foundations
	3 The SafeOps Approach
		3.1 Prerequisites
		3.2 Safety Engineering Support
		3.3 Safety Management Dashboard
		3.4 Data-Driven Safety
	4 Related Work
	5 Conclusion and Future Work
	References
Dynamic Checks of Evidence Models for Assurance Projects in Eclipse OpenCert
	1 Introduction
	2 Approach
		2.1 Extensible Evidence Checks
		2.2 Dynamic Evidence Checks and Log Streaming
	3 Related Work
	4 Conclusion and Perspectives
	References
An Approach for Deriving Reduced Collision Scenarios for Highly Automated Driving Systems
	1 Introduction
	2 State-of-Art and Science
		2.1 Standards and Regulations
		2.2 Highly Automated Driving System
		2.3 Variance-Based Sensitivity Analysis
	3 Scenario Reduction Approaches
		3.1 Scenario Description
		3.2 Logical Scenario Reduction
		3.3 Scenario Simulation and Sensitivity Analysis
	4 Conclusion and Future Work
	References
State of the Art Study of the Safety Argumentation Frameworks for Automated Driving System
	1 The Need for a Safety Argumentation Framework
	2 A New Safety Assurance Scheme and Global Initiatives
		2.1 Regulatory Informal Working Groups for ADS Safety Topics
		2.2 Global Standardization Activities
		2.3 Research and Industry Communities, Initiatives, and Projects
	3 Safety Case as a Core Approach to Argue the Safety of ADS
		3.1 Safety Case Content and Format
		3.2 Safety Goals
		3.3 Acceptance Criteria
	4 Safety Evidence: An Attempt to “Quantify” the ADS Safety
		4.1 Requirements-Based Testing with Numerical Approaches
		4.2 Quantitative Hazard and Risk Analysis as a Part of the SOTIF Processes
	5 Conclusions and Outlook
	References
1st International Workshop on Security and Safety Interaction (SENSEI 2022)
En
	1 Introduction
	International Program Committee 2022
Increasing IoT Security by Supply Power Measurement
	1 Introduction
		1.1 Attack Scenarios
		1.2 Pre-requisites and Assumptions
		1.3 Other Approaches/State of the Art
	2 Medical IoT Devices
	3 Current Measurement for Security
		3.1 Description of the Expected Power Consumption
		3.2 Technical Realization with Off-the-Shelf Hardware
		3.3 Extension to Detect Replay Attacks
	4 Results
	5 Conclusion
	References
Towards Interdependent Safety Security Assessments Using Bowties
	1 Introduction
	2 Diagrammatic Representations
	3 Formalisation of Safety-Security Bowties
		3.1 Structure Trees
		3.2 Disruption Trees
		3.3 Prevention Assessments
		3.4 Consequence Assessments
		3.5 Bowties
	4 Interdependence Analysis
		4.1 Running Example: Safety Security Interplay
		4.2 Independence
		4.3 Conditional Dependency
		4.4 Reinforcing Dependency
		4.5 Antagonistic Dependency
	5 Conclusions
		5.1 Related Work
		5.2 Assessment Techniques
		5.3 Next Steps
	References
3rd International Workshop on Underpinnings for Safe Distributed Artificial Intelligence (USDAI 2022)
En
	1 Introduction
	2 This Year’s Workshop
		Acknowledgement
	Workshop Committees
		Organization Committee
		Programme Committee
On Explainability in AI-Solutions: A Cross-Domain Survey
	1 Introduction
	2 Related Work
	3 What Is Explainability of AI
	4 Domains for Expainability
	5 Conclusion
	References
Research Questions in the Acceptance of Cybersecurity by SMEs in the EU
	1 Introduction
	2 Status Quo for Security Practices in SMEs
		2.1 Areas Addressed by SCRATCh Project
		2.2 Developments Within the SCRATCh Project
		2.3 The Problems that SMEs Face Regarding Security
	3 Holistic Angle
		3.1 Open Questions
	4 Conclusions
5th International Workshop on Artificial Intelligence Safety Engineering (WAISE 2022)
En
	1 Introduction
	2 Programme
	Session 1: AI Safety
	Session 2: Advances in Safe AI (Invited Talks 1)
	Session 3: Advances in Safe AI (Invited Talks 2)
	Session 4: Assurances for Autonomous Systems
	3 Acknowledgements
Safe Design of Stable Neural Networks for Fault Detection in Small UAVs
	1 Introduction
	2 Related Works
		2.1 Fault Detection and Diagnosis(FDD)
		2.2 Stability Quantification Through Lipschitz Constant
	3 Spectral Normalization Control of the Model Stability
	4 Validation on Drone Flights Data
		4.1 Dataset Description
		4.2 Binary Classification Results
		4.3 Multi-class Classification Results
	5 Conclusion
	References
Neural Criticality Metric for Object Detection Deep Neural Networks
	1 Introduction and Motivation
	2 Background and Related Work
		2.1 Overview of Object Detection Approaches
		2.2 Functional Safety, Verification and Validation
	3 From Classification to Object Detection Neural Criticality
		3.1 Formal Definition of the NC Metric
	4 Evaluation
		4.1 Experiment Setup
		4.2 Results and Discussion
		4.3 Models\' Architectural Details
	5 Concluding Remarks
	References
StaDRe and StaDRo: Reliability and Robustness Estimation of ML-Based Forecasting Using Statistical Distance Measures
	1 Introduction
	2 Related Work
		2.1 Dataset Shift
		2.2 Reliability
		2.3 Robustness
	3 Problem Definition
	4 Proposed Method
		4.1 SDD Vs. Performance
		4.2 Statistical-Distance-Based Reliability Estimate - StaDRe
		4.3 Statistical-Distance-Based Robustness - StaDRo
	5 Experiments
	6 Results and Discussion
	7 Conclusion and Future Works
	References
A Safety Assurable Human-Inspired Perception Architecture
	1 Introduction
	2 How Humans (Probably) Do Classification
	3 An Assurable Human-Inspired Classification Architecture
	4 Validation
	5 Conclusion
	References
Object Detection with Probabilistic Guarantees: A Conformal Prediction Approach
	1 Introduction
	2 Background: Conformal Prediction
	3 Experimental Setting and Goals
	4 Coordinate-Wise and Box-Wise Conformalization
		4.1 Preliminary Assignment
		4.2 Coordinate-Wise Conformalization
		4.3 Bonferroni Correction for Box-Level Guarantees
	5 Image-Wise Conformalization
	6 Statistical Pitfalls
	7 Discussions and Future Works
	References
Interval Weight-Based Abstraction for Neural Network Verification
	1 Introduction
	2 Background
		2.1 Neural Networks
		2.2 Verification of Neural Networks
	3 Proposed Approach
		3.1 Main Idea
		3.2 Abstraction for NNs with Tanh
		3.3 Abstraction for NNs with Relu
	4 Early Experiments
	5 Conclusion
	References
Logically Sound Arguments for the Effectiveness of ML Safety Measures
	1 Introduction
	2 Addressing Insufficiencies in Object Detection
	3 From Structural to Logical Argumentation
	References
Author Index