• Energy Research

Recent advances in artificial intelligence (AI) capabilities have increased the eagerness of companies to integrate AI into software systems. While AI can be used to have a positive impact on several dimensions of sustainability, this is often overshadowed by its potential negative influence. While many studies have explored sustainability factors in isolation, there is insufficient holistic coverage of potential sustainability benefits or costs that practitioners need to consider during decision-making for AI adoption. We therefore aim to synthesize trade-offs related to sustainability in the context of integrating AI into software systems. We want to make the sustainability benefits and costs of integrating AI more transparent and accessible for practitioners. The study was conducted in collaboration with a Dutch financial organization. We first performed a rapid review that led to the inclusion of 151 research papers. Afterward, we conducted six semi-structured interviews to enrich the data with industry perspectives. The combined results showcase the potential sustainability benefits and costs of integrating AI. The labels synthesized from the review regarding potential sustainability benefits were clustered into 16 themes, with "energy management" being the most frequently mentioned one. 11 themes were identified in the interviews, with the top mentioned theme being "employee wellbeing". Regarding sustainability costs, the review discovered seven themes, with "deployment issues" being the most popular one, followed by "ethics & society". "Environmental issues" was the top theme from the interviews. Our results provide valuable insights to organizations and practitioners for understanding the potential sustainability implications of adopting AI. Accepted for publication at the 8th International Workshop on Green and Sustainable Software (GREENS'24), collocated with ICSA'24

The onslaught of artificial intelligence (AI) in the global scientific and industrial landscape has brought with it far-reaching implications into how the software development process can be transformed. This article presents a systematic literature review focused on the integration of AI into the software development life cycle (SDLC) with a specific emphasis on sustainability. The research explores the application of AI in all facets of the SDLC. To this end, we structure 34 primary studies into the different stages of the SDLC, including requirements elicitation, analysis/design, development, testing, and deployment, while considering multiple dimensions of sustainability. Our findings present a synthesis of commonly used AI approaches under various aspects. These encompass guidelines for (i) automating requirements formulation, (ii) designing sustainable software, (iii) enhancing energy efficiency and code reuse, and (iv) effectively testing software. Environmental sustainability was found to be the most common dimension in the literature, primarily addressing energy efficiency and electronic waste. Additionally, we identify gaps in the literature, particularly the absence of addressing AI in SDLC from the angle of social sustainability and the lack of integration into developer toolkits.

In the daily work of a software architect, knowledge is ubiquitous. In addition to technical expertise, architecture knowledge (AK) also requires practical experience in the representation, communication, and management of architectural decisions. However, there is a pressing need to also incorporate sustainability aspects, i.e., capturing decisions towards software systems that are environmentally, economically, and socially balanced in the long term.With this study, we aim to provide a review of AK concepts and their representation and communication from a practical point of view. Having this understanding, we explore where sustainability can be applied in daily practice and how we can address sustainability in architecture processes in the future.The paper presents an empirical study conducted in an industrial context encompassing a questionnaire survey with 32 participants and semi-structured interviews with 15 practitioners; both groups are from a major bank in the Netherlands.Based on the insights gained from combining our findings, we (i) provide a map of applied concepts for communicating and representing AK in a large enterprise, and (ii) discuss potential avenues for carving sustainability into current software architecture practice.

Integrating (and evaluating) energy efficiency tactics into daily industrial practice is challenging. This paper addresses the experimental evaluation of energy efficiency tactics in industrial contexts. Based on different real-world scenarios, we assess five energy efficiency tactics for cloud-based software through individual experiments conducted across two companies. The results of the experiments show significant improvements in energy efficiency for three tactics, with two others showing enhanced efficiency albeit without statistical significance. In addition to the experiments, we draw lessons learned and practical insights into utilizing tactics in industrial contexts. Our results could guide practitioners in selecting and applying the most suitable tactic for their individual context. By linking tactics that emerged in the literature with evidence-based measures, we help including sustainability in software architecture design decision making.

The COVID-19 pandemic forced researchers to move academic conferences to a virtual format; but also brought attention back to the carbon footprint of their physical format. In general, while conferences can follow different formats with a different carbon footprint, the related factors of influence remain unclear, hence hindering informed decisions on how to organize and attend them.This work provides a preliminary study of the carbon footprint of academic conferences and the trade-offs between alternative conference types. First, we conducted a systematic literature review (SLR) to identify factors that contribute to the carbon footprint of on-site, virtual, and hybrid conferences. Second, we conducted an interview survey among steering committee members of a pilot of prominent international conferences to complement the SLR.There is agreement in the literature and the research community that on-site conferences suffer from travel-related emissions among many other factors. While the on-site type benefits from strong networking possibilities, the virtual and hybrid types can reduce carbon emissions significantly. Notwithstanding, we miss a generic framework that accounts for all revealed carbon footprint factors in each conference type. Also, compared to carbon offsetting, carbon handprinting as a footprint reduction option is considered in neither the literature nor the research community. Among the results, we provide a first sustainability model to compare current and future conference types according to their sustainability trade-offs. The model can be used as a decision-making tool by, e.g., conference organizers.

Including sustainability concerns in software architecture design decision-making is becoming imperative. In this hands-on tutorial, we will explain the notion of software sustainability and how software architecture can be pivotal to design for sustainability, setup up metrics and measures, and ultimately create a incremental process of design, measurement, and monitoring, and reassess past decisions so that the target sustainability goals can be achieved. To this aim, the tutorial will teach the sustainability assessment framework (SAF) toolkit, an open source toolkit for software sustainability design and sustainability-quality assessment.