Using Sustainability Impact Scores for Software Architecture Evaluation

TL;DR

The paper tackles the challenge of quantifying sustainability in software architecture by extending the Sustainability Impact Score (SIS) with a risk- and importance-based prioritization framework and cross-dimension analysis across Economic, Environmental, Social, and Technical dimensions. It introduces cross-dimension SIS definitions, a Utility Matrix-inspired prioritization, and normalization to enable objective comparisons, grounded in Dependency Matrices and Decision Maps. The approach is validated in an industrial energy-domain case (MMvIB), revealing that technical quality concerns can drive cross-dimensional sustainability effects and that multi-model architectures can yield higher sustainability support in certain dimensions, relative to a theoretical optimum. Collectively, the work provides a practical method for early, regulator-aligned SA decisions and a foundation for benchmarking sustainability trade-offs in digital solutions, with future work targeting multi-scenario and multi-stakeholder analyses along with benchmarking.

Abstract

For future regulatory compliance, organizations must assess and report on the state of sustainability in terms of its impacts over time. Sustainability, being a multidimensional concern, is complex to quantify. This complexity further increases with the interdependencies of the quality concerns across different sustainability dimensions. The research literature lacks a holistic way to evaluate sustainability at the software architecture level. With this study, our aim is to identify quality attribute (QA) trade-offs at the software architecture level and quantify the related sustainability impact. To this aim we present an improved version of the Sustainability Impact Score (SIS), building on our previous work. The SIS facilitates the identification and quantification of trade-offs in terms of their sustainability impact, leveraging a risk- and importance-based prioritization mechanism. To evaluate our approach, we apply it to an industrial case study involving a multi-model framework for integrated decision-making in the energy sector. Our study reveals that technical quality concerns have significant, often unrecognized impacts across sustainability dimensions. The SIS coupled with QA trade-offs can help practitioners make informed decisions that align with their sustainability goals. Early evaluations can help organizations mitigate sustainability risks by taking preventive actions.

Figures (3)

• Figure 1: DMatrix Overview
• Figure 2: DMatrices for a. Serverless (left) b. Containerization (middle) c. Theoretical Optimal (right)
• Figure 3: Decision Map showing quality concerns of the MMvIB IT Architecture using Multi Model Approach