Biologically inspired autonomous and autonomic systems (AAS) are essentially about creating self-directed and self-managing systems based on metaphors such as that of the autonomic nervous system. Agent technologies have been identified as a key enabler for engineering autonomy and autonomicity in systems, both in terms of retrofitting into legacy systems and designing new systems. Handing over responsibility to the systems raises concerns for humans. This invention is a result of continued investigation into a strand of research on how to engineer self-protection mechanisms into the systems to assist in providing mechanisms to control the scope and consequences of using autonomy and autonomicity.
This invention builds on previous mechanisms: using the apoptosis metaphor to potentially provide a self-destruct signal between autonomic agents when needed, and an Alice signal to facilitate self-identification and self-certification between anonymous autonomous agents and systems. The new approach is autonomic quiescence, where, for instance, an agent or craft has a self-sleep (quiescent state) built in, much the same as the mechanism for self-destruction. This would be used to protect the individual or system/mission goals, but be less extreme than the apoptosis/self-destruct mechanism, as an agent or craft may be brought out of a quiescent state if the concerns are resolved.
Autonomic computing is dependent on many disciplines for its success — not the least of these is research in agent technologies. At this stage, there are no assumptions that agents have to be used in an autonomic architecture, but as in complex systems, there are arguments for designing the system with agents, as well as providing in-built redundancy and greater robustness through to retrofitting legacy systems with autonomic capabilities that may benefit from an agent approach. Emerging research suggests that the autonomic manager may be an agent itself; for instance, an agent termed a self-managing cell (SMC), containing functionality for measurement and event correlation, and support for policy-based control.
To facilitate the aim of creating robust, dependable, self-managing systems, fault-tolerant mechanisms such as a heartbeat monitor (‘I am alive’ signals) and pulse monitor (urgency/reflex signals) may be included within the autonomic element. The notion behind the pulse may be included within the autonomic element. The notion behind the pulse monitor (PBM) is to provide an early warning of a condition so that preparations can be made to handle the processing load of diagnosis and planning a response, including diversion of load. Together with other forms of communication, it creates dynamics of autonomic responses — the introduction of multiple loops of control, some slow and precise, others fast and possibly imprecise, fitting with the biological metaphor of reflex and healing.
This work was done by Michael Hinchey of Goddard Space Flight Center and Roy Sterritt of University of Ulster, Northern Ireland. GSC-15176-1