Past Events › SERC

The Autonomy and Navigation Technology (ANT) Center at the Air Force Institute of Technology (AFIT) has begun development of a new Reference Architecture for Autonomy to begin to answer the myriad of needs within research, development and testing of autonomous and/or cooperative systems. This Reference Architecture will initially cater to small unmanned aerial and ground vehicles and the interfaces and operators that would be used to direct these vehicles in accomplishing tasks. The framework should be flexible to allow achievement of effective mission capabilities with autonomous and cooperative agents.

The Naval Postgraduate School and Air Force Institute of Technology have developed affordability tradeoff analysis methods employing architectural and parametric cost models applied to Unmanned Aerial Systems (UAS). A focus is at the interface of cost and architecture models for the automatic extraction of system size attributes from the architecture models at the levels of software, system and UAS domain product line. Costs can then be associated with architecture variants to assess tradeoffs between affordability and other ilities.
Cost models have been extended for integration with formal Model-Based Systems Engineering (MBSE) methods including SysML, Orthogonal Variability Modeling (OVM), discrete event simulation, and system behavioral modeling with Monterey Phoenix (MP). Relevant case studies for small UAS are continuing as the methods are being refined. Existing models were applied to demonstrate tradeoff analyses, and new models were created to better represent actual conditions empirically observed in practice.
The talk will overview the modeling methods and their interfaces, case studies applied to single system architectures, and current investigations with UAS domain reference architectures developed by AFIT.

One of the primary inhibitors of trust in autonomous and cyber-physical systems is the likelihood that any such computer-based systems may be infiltrated and manipulated by a malicious entity. We can systems engineer trust into such systems if we start with an in-depth understanding of the mathematical structure of the hardware and software. We then can migrate the computing from register-based CPUs with operating systems to pure dataflow based networked computing. Internal self-knowledge and constraints then can be trusted to limit autonomy and to protect cyber-physical systems from operating outside of the specified physical envelope. This talk will review the mathematics and provide a perspective of networked computing for the emerging discipline of cyber-physical systems engineering.

This presentation will provide a history of SERC Mission Aware research, current work to standardize assurance design in MBSE tools, and a discussion of opportunities for future research. The presentation will include a demonstration of a mission aware meta-model for an example complex system-of system.