An Initial Ontology for System Qualities

Abstract

This paper presents an initial ontology for reasoning about a system's System Qualities (SQs), ilities, or non‐functional requirements (reliability, usability, affordability, and more). The need for such ontology is based primarily on two factors. One is the importance of getting the SQs sufficiently well defined such that the system's definition, development, and evolution result in a satisfactory balance of SQ values for the system's success‐critical stakeholders, given the frequent system shortfalls and overruns that occur when the system does not achieve this balance. The other is that current system acquisition and evolution guidance descriptions have numerous deficiencies and inconsistencies in their coverage of SQ considerations. This situation is becoming more serious as systems and their stakeholders become increasingly complex, dynamic, and diverse. This paper provides an elaboration of the needs, a set of initial SQ ontology elements and definitions, examples of their application to some key SQs and their relationships, and an identification of further research and development needed to make the ontology fully useful and evolvable.

Balancing Systems Engineering Rigor with Agile Software Development Flexibility

“…This helped everyone in the project better understand the product, their individual role in bringing the product to life, and where in the lifecycle that product currently resided. This understanding included not only the functional aspects of the product but also the non-functional aspects which support the quality attributes of that product (Boehm and Kukreja 2017). Further, the maturity expectations (or, more precisely in this case, the expectations of product immaturity) helped define phases where the agile developers could have extensive freedom in terms of design flexibility and creativity.…”

Systemic virtues as a foundation for a general theory of design elegance

“…Many further examples of such concepts have been discussed in the engineering research literature, for example resilience, sustainability, adaptiveness, recyclability, maintainability, upgradability, manufacturability, affordability, scalability, survivability, versatility, interoperability, modularity, flexibility, and agility (Boehm & Kukreja, ; Madni, ; Adam M. Ross & Rhodes, ; Watson & Griffin, ; Watson et al, ). We note that most of these attibutes are subsumed under the general rubric of ilities, but that this may be due to insuffiencient attention being paid to other kinds of systemic virtues.…”

“…This was immediately problematic, as the terms as used in practice do not have consistent definitions. This issue is well known and has been a subject of research in recent times (Boehm & Kukreja, ; Dou, Wang, Tang, Ross, & Sullivan, ; Adam M. Ross & Rhodes, , ). Our view is that progress in this area will be crucial for building a theory of design elegance that strongly exhibits the theoretical virtues.…”

“…Intuitively, and by analogy with the value of other classes of virtues, it seems reasonable to suppose that systems that more comprehensively exhibit the systemic virtues are more likely to maintain their viability, succeed in intended missions, be useful in a variety of unanticipated contexts, and be platforms for upgradability to more highly performant systems. Limited formal studies have been done in this area, but initial work confirms this assessment (Boehm & Kukreja, 2015;de Weck, Ross, & Rhodes, 2012;Middlebrooks, Rhodes, Cipolloni, & Goerger, 2019;A. M. Ross, 2006).…”

“…The classification project showed that dependencies between the virtues could be complex, with many linkages between them, and that these links might be of different kinds, for example, some virtues may reinforce each other, and some may oppose each other. This kind of dynamics has been explored by others, for example, using a matrix view to expose synergies and conflicts (Boehm & Kukreja, 2015). We developed our own Nsquared diagram, a 2 × 2 matrix having on each axis our list of sample virtues, and then noted in each cell whether a change to the virtue labelled in its row resulted in a change in the virtue labelled in the corresponding column.…”