https://www.mdu.se/

Integrating safety and security in automotive cyber-physical systems (CPS) domains (e.g. autonomous vehicles), is challenging for two main reasons. First, it is still difficult to represent the potential consequences of system failures or malicious attacks. Secondly, these systems must ensure safety and security despite unknowns and uncertainties. A Digital Dependability Identity (DDI) can facilitate this by encapsulating all dependability characteristics (e.g., design, requirements, safety, and security analysis models) of CPS's components. The Open Dependability Exchange (ODE) metamodel is an implementation of the DDI concept, but has limitations in the interplay between safety and security. ODE is aligned with with ISO 26262 but lacks certain security concepts to be aligned with ISO 21434. Also, ODE supports modeling fault trees, but modeling attack trees and attack-fault trees still not. This paper proposes an extension to the ODE metamodel, aiming to increase coverage of ISO 21434 concepts and allowing the modeling of attack-fault trees. We built these metamodel extensions based on an analysis of the ODE metamodel, industry standards, Microsoft STRIDE model, and HEAVENS security analysis methodologies. We evaluated the proposed extensions in an illustrative example of an autonomous vehicle.

To win the competition, diversification via software-implemented functionality, has become a trend in various do-mains. In the automotive domain, for instance, road vehicles are being transformed from steels and wheels to software and services, where software is expected to shape the vehicles virtuously, over their entire lifetime, by increasing customer’s satisfaction via an increasing number of comfort features. However, if safety as-surance is not in focus while managing variability, these features, may viciously turn into risk-bearing features potentially leading to fatalities. In the literature, the problem of safety assurance and its configuration has been addressed. The usage of the Base Variability Resolution language, for instance, was explored to systematise and configure not only product-related features but also process and assurance-related features, necessary for the justification purposes. The problem of safety assurance has also been addressed, yet without variability aspects, from a socio-technical perspective, e.g., by Rasmussen with his socio-technical system. In this position paper, we propose an extension of th Rasmussen’s socio-technical system. Our extension is twofold: on one hand it embraces product lines instead of single systems,on the other hand it develops the technical and argumentation aspects in addition to the socio-aspects. We also mention how our extension can be specialised in the automotive context.

Compliance management is a challenging activity in regulated industries. This is due to the need of navigating rapidly shifting requirements. In the automotive domain products (items) evolve rapidly creating a product line in time in addition to the one in space. The product variability is constrained by legislation, standards, etc. whose applicability may vary depending on the jurisdiction. In this paper, we focus on the legislation and its impact on the product in terms of inclusion/exclusion of product features. We exploit the Semantic Web technologies and we propose an ontology-based representation for managing product variability in compliance with legislation. Specifically, we provide the representation in Resource Description Framework (RDF) and we introduce reusable SHACL (Shapes Constraint Language) constraints to shape the RDF graphs. We illustrate our proposal by considering US and UNECE regulations and their impact on the window lifter.

Technological changes such as the use of Augmented Reality (AR) along with the advent of new organizational changes such as digitalization are on the one hand positively changing the way of working but on the other hand they are introducing new risks, potentially leading to not only normal but also post-normal accidents. In our previous work, we have incrementally proposed a novel framework, called FRAAR, for risk assessment of AR-equipped socio-technical systems (i.e., systems integrating human, organizational and technical entities (such as AR)). We have also partly evaluated our framework via an industrial automotive study and by providing comparison and positioning with respect to other related works in a systematic literature review. In this paper, we conduct a new study to evaluate the applicability and effectiveness of our framework in a different domain. To do that, we choose a digitalized socio-technical factory system, focusing on the human–robot collaboration for a realistic diesel engine assembly task using AR-based user interface in an organization affected by organizational changes. Then, we design and execute our study to apply our framework and we discuss about the extent the conceptualizations provided by the framework are effective to capture the essential information for risk assessment in socio-technical robotic manufacturing, the extent the robotic safety standards are supported (to demonstrate the applicability of the framework in the robotic domain) and the extent of effectiveness of the risk assessment with respect to AR and organizational changes. Finally, we discuss about validity of our work and we provide our findings and intended future work.

This book constitutes the proceedings of the Workshops held in conjunction with the 43rd International Conference on Computer Safety, Reliability, and Security, SAFECOMP 2024, which took place in Florence, Italy, during September 2024. 

The 36 papers included in this book were carefully reviewed and selected from a total of 64 submissions to the following workshops: 

DECSoS 2024 – 19th Workshop on Dependable Smart Embedded and Cyber-Physical Systems and Systems-of-SystemsSASSUR 2024 - 11th International Workshop on Next Generation of System Assurance Approaches for Critical SystemsTOASTS 2024 – Towards A Safer Systems’ Architecture Through SecurityWAISE 2024 – 7th International Workshop on Artificial Intelligence Safety Engineering

Next-generation human-computer interactions comprise the interactions that are expected to take place between e.g., pedestrians and the so-called External Human-Machine Interfaces (EHMIs). EHMIs are expected to replace the in vehicle Driver-oriented Driver-Machine Interface in the context of high automation levels, where the driving task is fully automated. In this context, it is paramount to ensure absence of unreasonable risk due to hazards resulting from functional insufficiencies of the intended functionality or its implementation, i.e., Safety Of The Intended Functionality (SOTIF), defined in ISO 21448:2022. ISO 21448:2022 has been applied to ensure the Safety Of The Intended Driver-Machine Interface-Functionality. SOTIF does not include any specific consideration to address the so-called External Human-Machine Interface. Hence, in this position paper, we posit that the functional insufficiencies shall also embrace the socio-technical issues that might emerge in the extended automated driving system's ecosystem. These issues include faulty interactions with pedestrians, traffic-lights, and more broadly, by taking a vehicle to everything perspective (VtoX), with everything.