Systems Engineering and Architecting for Intelligent Autonomous Systems

Methodology and Architecture for Safety Management

“…Several run-time models that have been experienced and surveyed to cope with uncertainty have been proposed [11,12]. They are presented as an appropriate solution to cope with the concerns of safety in AV emphasised previously.…”

“…Regarding software architecture, modularization and abstraction have been the first stages to break down the complexity of systems in programs and architectures. However, according to Dragoni et al [14], those stages still result into the creation of monoliths, where monolithic applications are generally built as a single unit with functionalities spread in modules and components that are not independently executable. The final product and code result in difficulties for scalability, maintainability and evolvability.…”

“…The approach suggests to isolate all business capabilities of the systems into simple and small specialized services similar to components but decoupled enough to be independently deployable by an automated deployment machinery. A micro-service results to be an independent functional process that express one capability, also defined as cohesive [14], interacts with messages and owns its domain logic locally. [16] is an application of micro-service architecture for aerial unmanned vehicle.…”

Challenges for the Self-Safety in Autonomous Vehicles

“…This work has introduced practical methods to deal with context with sets of layers of abstraction in the same functional architecture. On this basis, further works have proposed solutions to solve part of the complexity of the domain by identifying the degradation states for well-adapted decisions and actions [4,5], apprehending the range and limitation of actions of the system [6,7], evolving the architecture of the system to provide better adaptation behaviors, and finally, integrating non-functional dimensions [8].…”

“…However, with the advent of automated vehicles that requires more and more understanding of its context, the necessity to build an extendable, context-aware, evolvable and pluggable architecture has arisen at run-time [9]. In fact, levels 4-5 ADS-operated vehicles require a well-adapted architecture aimed at developing the appropriate functions, capabilities and skills to adapt to complex and dynamic environment situations while still guaranteeing non-functional properties, as safety in any situations [8,10]. Besides, the surge of microservices paradigm [11] and the Internet of Things technologies [12,13] are providing new semantic and abstraction tools for the knowledge representation and architecture management for the field.…”