Multi-agent architectures for high assurance systems

High assurance encompasses availability, security, safety and any other features that affect the predictability of the system. Techniques from disciplines like fault tolerant computing, real time systems and reliability engineering are applied to improve one or many of the features listed above. This ad hoc approach severely limits the systems capabilities and often leads to brittle systems. High assurance systems engineering attempts to adopt a holistic approach to the problem of designing mission critical systems. Multi-agent architectures are a new paradigm in conceptualizing, designing and implementing systems. Agent architectures are characterized by the concepts of situatedness, autonomy and flexibility. These characteristics are essential for building high assurance mission critical systems. In this paper, we propose a multi-agent based approach for designing high assurance systems.