Situation awareness is key for many domains. Autonomous driving, air traffic control, maritime safety and security, emergency control, factory and plant surveillance, but also sustainability via energy saving with smart buildings or even cities – they all require being aware of the environmental situation.
Collecting and interpreting large amounts of information coming from many sources with various levels of reliability and trust is thus the initial step for many organizations in a sense – think – act chain that allows them to make informed decisions and take appropriate actions.
Interwoven with our Digital Twin research and aiming at advanced smart systems and resilient infrastructures, ESI supports this with the development of information-centric architectures for systems-of-systems.
In many domains, a situation awareness application will consist of a coalition of autonomous systems. This is visible, e.g., in the maritime domain, where stakeholders like the Coast Guard rely on systems-of-systems that can evolve dynamically and adapt to new sources.
As we show in the figure below, the information flow of such an application consists of several steps for various information streams with quality checks, subsequent alignment and fusion that uses semantic reasoning and AI-based recognition of objects of interest and finally of the situation at hand.
Automated alerts, predictions, and state-of-the-art visualizations help human operators in their decision-making. Moreover, the system also uses the situation awareness while it protects sensitive information, e.g., by relaxing policies if an emergency requires this.
We develop such an information-centric architecture using a user-centric 5-steps approach:
Identify tasks and views of system operators and users
Capture required information sources, processing, and reasoning
Analyze the resulting information flows and their behaviors over time, which also identifies non-functional requirements
Determine information flow mechanics for core data and meta‐data using channels, transport, feedback, and storage mechanisms
Create a composition of functionality and flow mechanics
Seeing this as a complex architecture task, we find the A3 Method helps here to develop, consolidate, and communicate central aspects.
Furthermore, we support the design of the necessary behavior and information flows (step 3) with a light-weight modelling and analysis technique:
Our Probabilistic System Summaries for Behavior Architecting allow to explore design choices like where to put the ‘smarts’ of a smart system, or, more general, offers causal analysis of the costs and benefits of information-centric decision strategies in key scenarios. This lets us bridge between two essential aspects: first, the architectural point which information is available and processed where and when in the system, and second the effects of that information processing and decision making given the expected environment and tasks.