Towards an Engineering Discipline for Resilient Cyber-Physical Systems

TL;DR

This paper advocates for a new engineering discipline focused on designing resilient cyber-physical systems capable of adapting to environmental changes, emphasizing software tools for self-management and resilience.

Contribution

It introduces foundational concepts, methods, and tools for engineering resilient cyber-physical systems, supported by studies, software packages, and community resources.

Findings

Developed new methods and tools for resilience

Created software packages and guidance resources

Conducted studies to understand resilience challenges

Abstract

Resilient cyber-physical systems comprise computing systems able to continuously interact with the physical environment in which they operate, despite runtime errors. The term resilience refers to the ability to cope with unexpected inputs while delivering correct service. Examples of resilient computing systems are Google's PageRank and the Bubblesort algorithm. Engineering for resilient cyber-physical systems requires a paradigm shift, prioritizing adaptability to dynamic environments. Software as a tool for self-management is a key instrument for dealing with uncertainty and embedding resilience in these systems. Yet, software engineers encounter the ongoing challenge of ensuring resilience despite environmental dynamic change. My thesis aims to pioneer an engineering discipline for resilient cyber-physical systems. Over four years, we conducted studies, built methods and tools, delivered software packages, and a website offering guidance to practitioners. This paper provides a condensed overview of the problems tackled, our methodology, key contributions, and results highlights. Seeking feedback from the community, this paper serves both as preparation for the thesis defense and as insight into future research prospects.