Affiliations: Systems Architecting and Engineering Program,
University of Southern California, Los Angeles, CA, USA
Abstract: System resilience is the ability of organizational, hardware and
software systems to mitigate the severity and likelihood of failures or losses,
to adapt to changing conditions, and to respond appropriately after the fact.
Challenger, Columbia, Chernobyl and Bhopal are examples of such failures.
System resilience goes beyond traditional disciplines, such as reliability and
system safety to achieve its goal. System resilience requires that technical
and managerial processes be executed as a unified whole. System resilience
employs the systems approach at product and infrastructure levels. The
infrastructure system includes such nodes as the developer, the customer, the
user, the maintainer, and the operator. System resilience requires that systems
engineering principles be practiced across organizational boundaries and to a
greater level of detail than is common in today's world. System
resilience incorporates the principles of organizational psychology to develop
beneficial paradigms within all nodes of the infrastructure. The combination of
capabilities, culture and infrastructure forms the basic framework of system
resilience. A key aspect of catastrophes is emergence, that is, characteristics
of a system that cannot be predicted from the characteristics of the
components. Sometimes emergence results from the negative interaction among two
or more elements of the system, which, when acting alone, perform benignly.
While positive, or desirable, emergence is widely studied, negative, or
undesirable, emergence is of interest in the study of resilience. Prediction of
emergence and design of adaptive systems that will survive emergent events is a
major challenge of resilience. Adaptability, agility and robustness are the
characteristics of a system that allow it to survive a disruption. Although the
principles of adaptability, agility and robustness have been formulated and
shown to reflect the characteristics of systems that have survived disruption,
the actual implementation of them lie largely in the heuristics of systems
architecting.
