Luís Cruz

Haiyin Zhang, Luís Cruz, Arie van Deursen

The popularity of machine learning has wildly expanded in recent years. Machine learning techniques have been heatedly studied in academia and applied in the industry to create business value. However, there is a lack of guidelines for code quality in machine learning applications. In particular, code smells have rarely been studied in this domain. Although machine learning code is usually integrated as a small part of an overarching system, it usually plays an important role in its core functionality. Hence ensuring code quality is quintessential to avoid issues in the long run. This paper proposes and identifies a list of 22 machine learning-specific code smells collected from various sources, including papers, grey literature, GitHub commits, and Stack Overflow posts. We pinpoint each smell with a description of its context, potential issues in the long run, and proposed solutions. In addition, we link them to their respective pipeline stage and the evidence from both academic and grey literature. The code smell catalog helps data scientists and developers produce and maintain high-quality machine learning application code.

Santiago del Rey, Luís Cruz, Xavier Franch et al.

To raise awareness of the environmental impact of deep learning (DL), many studies estimate the energy use of DL systems. However, energy estimates during DL training often rely on unverified assumptions. This work addresses that gap by investigating how model architecture and training environment affect energy consumption. We train a variety of computer vision models and collect energy consumption and accuracy metrics to analyze their trade-offs across configurations. Our results show that selecting the right model-training environment combination can reduce training energy consumption by up to 80.68% with less than 2% loss in $F_1$ score. We find a significant interaction effect between model and training environment: energy efficiency improves when GPU computational power scales with model complexity. Moreover, we demonstrate that common estimation practices, such as using FLOPs or GPU TDP, fail to capture these dynamics and can lead to substantial errors. To address these shortcomings, we propose the Stable Training Epoch Projection (STEP) and the Pre-training Regression-based Estimation (PRE) methods. Across evaluations, our methods outperform existing tools by a factor of two or more in estimation accuracy.

Roberto Verdecchia, June Sallou, Luís Cruz

With the ever-growing adoption of AI-based systems, the carbon footprint of AI is no longer negligible. AI researchers and practitioners are therefore urged to hold themselves accountable for the carbon emissions of the AI models they design and use. This led in recent years to the appearance of researches tackling AI environmental sustainability, a field referred to as Green AI. Despite the rapid growth of interest in the topic, a comprehensive overview of Green AI research is to date still missing. To address this gap, in this paper, we present a systematic review of the Green AI literature. From the analysis of 98 primary studies, different patterns emerge. The topic experienced a considerable growth from 2020 onward. Most studies consider monitoring AI model footprint, tuning hyperparameters to improve model sustainability, or benchmarking models. A mix of position papers, observational studies, and solution papers are present. Most papers focus on the training phase, are algorithm-agnostic or study neural networks, and use image data. Laboratory experiments are the most common research strategy. Reported Green AI energy savings go up to 115%, with savings over 50% being rather common. Industrial parties are involved in Green AI studies, albeit most target academic readers. Green AI tool provisioning is scarce. As a conclusion, the Green AI research field results to have reached a considerable level of maturity. Therefore, from this review emerges that the time is suitable to adopt other Green AI research strategies, and port the numerous promising academic results to industrial practice.

Tim Yarally, Luís Cruz, Daniel Feitosa et al.

Modern AI practices all strive towards the same goal: better results. In the context of deep learning, the term "results" often refers to the achieved accuracy on a competitive problem set. In this paper, we adopt an idea from the emerging field of Green AI to consider energy consumption as a metric of equal importance to accuracy and to reduce any irrelevant tasks or energy usage. We examine the training stage of the deep learning pipeline from a sustainability perspective, through the study of hyperparameter tuning strategies and the model complexity, two factors vastly impacting the overall pipeline's energy consumption. First, we investigate the effectiveness of grid search, random search and Bayesian optimisation during hyperparameter tuning, and we find that Bayesian optimisation significantly dominates the other strategies. Furthermore, we analyse the architecture of convolutional neural networks with the energy consumption of three prominent layer types: convolutional, linear and ReLU layers. The results show that convolutional layers are the most computationally expensive by a strong margin. Additionally, we observe diminishing returns in accuracy for more energy-hungry models. The overall energy consumption of training can be halved by reducing the network complexity. In conclusion, we highlight innovative and promising energy-efficient practices for training deep learning models. To expand the application of Green AI, we advocate for a shift in the design of deep learning models, by considering the trade-off between energy efficiency and accuracy.

Roberto Verdecchia, Luís Cruz, June Sallou et al.

With the growing availability of large-scale datasets, and the popularization of affordable storage and computational capabilities, the energy consumed by AI is becoming a growing concern. To address this issue, in recent years, studies have focused on demonstrating how AI energy efficiency can be improved by tuning the model training strategy. Nevertheless, how modifications applied to datasets can impact the energy consumption of AI is still an open question. To fill this gap, in this exploratory study, we evaluate if data-centric approaches can be utilized to improve AI energy efficiency. To achieve our goal, we conduct an empirical experiment, executed by considering 6 different AI algorithms, a dataset comprising 5,574 data points, and two dataset modifications (number of data points and number of features). Our results show evidence that, by exclusively conducting modifications on datasets, energy consumption can be drastically reduced (up to 92.16%), often at the cost of a negligible or even absent accuracy decline. As additional introductory results, we demonstrate how, by exclusively changing the algorithm used, energy savings up to two orders of magnitude can be achieved. In conclusion, this exploratory investigation empirically demonstrates the importance of applying data-centric techniques to improve AI energy efficiency. Our results call for a research agenda that focuses on data-centric techniques, to further enable and democratize Green AI.

Jai Kannan, Scott Barnett, Luís Cruz et al.

Meeting the rise of industry demand to incorporate machine learning (ML) components into software systems requires interdisciplinary teams contributing to a shared code base. To maintain consistency, reduce defects and ensure maintainability, developers use code analysis tools to aid them in identifying defects and maintaining standards. With the inclusion of machine learning, tools must account for the cultural differences within the teams which manifests as multiple programming languages, and conflicting definitions and objectives. Existing tools fail to identify these cultural differences and are geared towards software engineering which reduces their adoption in ML projects. In our approach we attempt to resolve this problem by exploring the use of context which includes i) purpose of the source code, ii) technical domain, iii) problem domain, iv) team norms, v) operational environment, and vi) development lifecycle stage to provide contextualised error reporting for code analysis. To demonstrate our approach, we adapt Pylint as an example and apply a set of contextual transformations to the linting results based on the domain of individual project files under analysis. This allows for contextualised and meaningful error reporting for the end-user.

Tim Yarally, Luís Cruz, Daniel Feitosa et al.

The batch size is an essential parameter to tune during the development of new neural networks. Amongst other quality indicators, it has a large degree of influence on the model's accuracy, generalisability, training times and parallelisability. This fact is generally known and commonly studied. However, during the application phase of a deep learning model, when the model is utilised by an end-user for inference, we find that there is a disregard for the potential benefits of introducing a batch size. In this study, we examine the effect of input batching on the energy consumption and response times of five fully-trained neural networks for computer vision that were considered state-of-the-art at the time of their publication. The results suggest that batching has a significant effect on both of these metrics. Furthermore, we present a timeline of the energy efficiency and accuracy of neural networks over the past decade. We find that in general, energy consumption rises at a much steeper pace than accuracy and question the necessity of this evolution. Additionally, we highlight one particular network, ShuffleNetV2(2018), that achieved a competitive performance for its time while maintaining a much lower energy consumption. Nevertheless, we highlight that the results are model dependent.

Wander Siemers, June Sallou, Luís Cruz

Artificial intelligence is bringing ever new functionalities to the realm of mobile devices that are now considered essential (e.g., camera and voice assistants, recommender systems). Yet, operating artificial intelligence takes up a substantial amount of energy. However, artificial intelligence is also being used to enable more energy-efficient solutions for mobile systems. Hence, artificial intelligence has two faces in that regard, it is both a key enabler of desired (efficient) mobile functionalities and a major power draw on these devices, playing a part in both the solution and the problem. In this paper, we present a review of the literature of the past decade on the usage of artificial intelligence within the realm of green mobile computing. From the analysis of 34 papers, we highlight the emerging patterns and map the field into 13 main topics that are summarized in details. Our results showcase that the field is slowly increasing in the past years, more specifically, since 2019. Regarding the double impact AI has on the mobile energy consumption, the energy consumption of AI-based mobile systems is under-studied in comparison to the usage of AI for energy-efficient mobile computing, and we argue for more exploratory studies in that direction. We observe that although most studies are framed as solution papers (94%), the large majority do not make those solutions publicly available to the community. Moreover, we also show that most contributions are purely academic (28 out of 34 papers) and that we need to promote the involvement of the mobile software industry in this field.

Bart van Oort, Luís Cruz, Babak Loni et al.

Machine Learning (ML) projects incur novel challenges in their development and productionisation over traditional software applications, though established principles and best practices in ensuring the project's software quality still apply. While using static analysis to catch code smells has been shown to improve software quality attributes, it is only a small piece of the software quality puzzle, especially in the case of ML projects given their additional challenges and lower degree of Software Engineering (SE) experience in the data scientists that develop them. We introduce the novel concept of project smells which consider deficits in project management as a more holistic perspective on software quality in ML projects. An open-source static analysis tool mllint was also implemented to help detect and mitigate these. Our research evaluates this novel concept of project smells in the industrial context of ING, a global bank and large software- and data-intensive organisation. We also investigate the perceived importance of these project smells for proof-of-concept versus production-ready ML projects, as well as the perceived obstructions and benefits to using static analysis tools such as mllint. Our findings indicate a need for context-aware static analysis tools, that fit the needs of the project at its current stage of development, while requiring minimal configuration effort from the user.

Bart van Oort, Luís Cruz, Maurício Aniche et al.

Artificial Intelligence (AI) and Machine Learning (ML) are pervasive in the current computer science landscape. Yet, there still exists a lack of software engineering experience and best practices in this field. One such best practice, static code analysis, can be used to find code smells, i.e., (potential) defects in the source code, refactoring opportunities, and violations of common coding standards. Our research set out to discover the most prevalent code smells in ML projects. We gathered a dataset of 74 open-source ML projects, installed their dependencies and ran Pylint on them. This resulted in a top 20 of all detected code smells, per category. Manual analysis of these smells mainly showed that code duplication is widespread and that the PEP8 convention for identifier naming style may not always be applicable to ML code due to its resemblance with mathematical notation. More interestingly, however, we found several major obstructions to the maintainability and reproducibility of ML projects, primarily related to the dependency management of Python projects. We also found that Pylint cannot reliably check for correct usage of imported dependencies, including prominent ML libraries such as PyTorch.

Thomas Durieux, Rui Abreu, Martin Monperrus et al.

Travis CI handles automatically thousands of builds every day to, amongst other things, provide valuable feedback to thousands of open-source developers. In this paper, we investigate Travis CI to firstly understand who is using it, and when they start to use it. Secondly, we investigate how the developers use Travis CI and finally, how frequently the developers change the Travis CI configurations. We observed during our analysis that the main users of Travis CI are corporate users such as Microsoft. And the programming languages used in Travis CI by those users do not follow the same popularity trend than on GitHub, for example, Python is the most popular language on Travis CI, but it is only the third one on GitHub. We also observe that Travis CI is set up on average seven days after the creation of the repository and the jobs are still mainly used (60%) to run tests. And finally, we observe that 7.34% of the commits modify the Travis CI configuration. We share the biggest benchmark of Travis CI jobs (to our knowledge): it contains 35,793,144 jobs from 272,917 different GitHub projects.

François Bechet, Jérôme Maquoi, Luís Cruz et al.

Green software engineering is emerging as a crucial response to information technology's rising energy impact, especially in continuous development. However, there remain challenges in devising automated methods for identifying energy regressions across commits and their associated code change patterns. In particular, little effort has been put into automatically detecting regressions at the commit level by identifying statistically significant changes in energy consumption. In this paper, we introduce EnergyTrackr, an approach designed to detect energy regressions across multiple commits that can then be used to identify code anti-patterns potentially contributing to the increase of software energy consumption over time. We describe our empirical evaluation, including repository mining and source code analysis, made on 3,232 commits from three Java projects, and show the approach's ability to identify significant energy changes. We also highlight recurring anti-patterns such as missing early exits or costly dependency upgrades. We expect EnergyTrackr to assist developers in accurately monitoring energy regressions and improvements within their projects, identifying code anti-patterns, and helping them optimize their source code to reduce software energy consumption.

Nienke Nijkamp, June Sallou, Niels van der Heijden et al.

Integrating Artificial Intelligence (AI) into software systems has significantly enhanced their capabilities while escalating energy demands. Ensemble learning, combining predictions from multiple models to form a single prediction, intensifies this problem due to cumulative energy consumption. This paper presents a novel approach to model selection that addresses the challenge of balancing the accuracy of AI models with their energy consumption in a live AI ensemble system. We explore how reducing the number of models or improving the efficiency of model usage within an ensemble during inference can reduce energy demands without substantially sacrificing accuracy. This study introduces and evaluates two model selection strategies, Static and Dynamic, for optimizing ensemble learning systems performance while minimizing energy usage. Our results demonstrate that the Static strategy improves the F1 score beyond the baseline, reducing average energy usage from 100% from the full ensemble to 62%. The Dynamic strategy further enhances F1 scores, using on average 76% compared to 100% of the full ensemble. Moreover, we propose an approach that balances accuracy with resource consumption, significantly reducing energy usage without substantially impacting accuracy. This method decreased the average energy usage of the Static strategy from approximately 62% to 14%, and for the Dynamic strategy, from around 76% to 57%. Our field study of Green AI using an operational AI system developed by a large professional services provider shows the practical applicability of adopting energy-conscious model selection strategies in live production environments.

Luís Cruz, João Paulo Fernandes, Maja H. Kirkeby et al.

The environmental impact of Artificial Intelligence (AI)-enabled systems is increasing rapidly, and software engineering plays a critical role in developing sustainable solutions. The "Greening AI with Software Engineering" CECAM-Lorentz workshop (no. 1358, 2025) funded by the Centre Européen de Calcul Atomique et Moléculaire and the Lorentz Center, provided an interdisciplinary forum for 29 participants, from practitioners to academics, to share knowledge, ideas, practices, and current results dedicated to advancing green software and AI research. The workshop was held February 3-7, 2025, in Lausanne, Switzerland. Through keynotes, flash talks, and collaborative discussions, participants identified and prioritized key challenges for the field. These included energy assessment and standardization, benchmarking practices, sustainability-aware architectures, runtime adaptation, empirical methodologies, and education. This report presents a research agenda emerging from the workshop, outlining open research directions and practical recommendations to guide the development of environmentally sustainable AI-enabled systems rooted in software engineering principles.

Roberto Vergallo, Luís Cruz, Alessio Errico et al.

'Follow-the-Sun' (FtS) is a theoretical computational model aimed at minimizing the carbon footprint of computer workloads. It involves dynamically moving workloads to regions with cleaner energy sources as demand increases and energy production relies more on fossil fuels. With the significant power consumption of Artificial Intelligence (AI) being a subject of extensive debate, FtS is proposed as a strategy to mitigate the carbon footprint of training AI models. However, the literature lacks scientific evidence on the advantages of FtS to mitigate the carbon footprint of AI workloads. In this paper, we present the results of an experiment conducted in a partial synthetic scenario to address this research gap. We benchmarked four AI algorithms in the anomaly detection domain and measured the differences in carbon emissions in four cases: no strategy, FtS, and two strategies previously introduced in the state of the art, namely Flexible Start and Pause and Resume. To conduct our experiment, we utilized historical carbon intensity data from the year 2021 for seven European cities. Our results demonstrate that the FtS strategy not only achieves average reductions of up to 14.6% in carbon emissions (with peaks of 16.3%) but also helps in preserving the time needed for training.

Luís Cruz, Xavier Franch Gutierrez, Silverio Martínez-Fernández

The latest advancements in machine learning, specifically in foundation models, are revolutionizing the frontiers of existing software engineering (SE) processes. This is a bi-directional phenomona, where 1) software systems are now challenged to provide AI-enabled features to their users, and 2) AI is used to automate tasks within the software development lifecycle. In an era where sustainability is a pressing societal concern, our community needs to adopt a long-term plan enabling a conscious transformation that aligns with environmental sustainability values. In this paper, we reflect on the impact of adopting environmentally friendly practices to create AI-enabled software systems and make considerations on the environmental impact of using foundation models for software development.

Rui Abreu, Marco Couto, Luís Cruz et al.

This report describes the research goals and results of the Green Software Lab (GSL) research project. This was a project funded by Fundação para a Ciência e a Tecnologia (FCT) -- the Portuguese research foundation -- under reference POCI-01-0145-FEDER-016718, that ran from January 2016 till July 2020. This report includes the complete document reporting the results achieved during the project execution, which was submitted to FCT for evaluation on July 2020. It describes the goals of the project, and the different research tasks presenting the deliverables of each of them. It also presents the management and result dissemination work performed during the project's execution. The document includes also a self assessment of the achieved results, and a complete list of scientific publications describing the contributions of the project. Finally, this document includes the FCT evaluation report.

Yuanhao Xie, Luís Cruz, Petra Heck et al.

The development of artificial intelligence (AI) has made various industries eager to explore the benefits of AI. There is an increasing amount of research surrounding AI, most of which is centred on the development of new AI algorithms and techniques. However, the advent of AI is bringing an increasing set of practical problems related to AI model lifecycle management that need to be investigated. We address this gap by conducting a systematic mapping study on the lifecycle of AI model. Through quantitative research, we provide an overview of the field, identify research opportunities, and provide suggestions for future research. Our study yields 405 publications published from 2005 to 2020, mapped in 5 different main research topics, and 31 sub-topics. We observe that only a minority of publications focus on data management and model production problems, and that more studies should address the AI lifecycle from a holistic perspective.

Mark Haakman, Luís Cruz, Hennie Huijgens et al.

Tech-leading organizations are embracing the forthcoming artificial intelligence revolution. Intelligent systems are replacing and cooperating with traditional software components. Thus, the same development processes and standards in software engineering ought to be complied in artificial intelligence systems. This study aims to understand the processes by which artificial intelligence-based systems are developed and how state-of-the-art lifecycle models fit the current needs of the industry. We conducted an exploratory case study at ING, a global bank with a strong European base. We interviewed 17 people with different roles and from different departments within the organization. We have found that the following stages have been overlooked by previous lifecycle models: data collection, feasibility study, documentation, model monitoring, and model risk assessment. Our work shows that the real challenges of applying Machine Learning go much beyond sophisticated learning algorithms - more focus is needed on the entire lifecycle. In particular, regardless of the existing development tools for Machine Learning, we observe that they are still not meeting the particularities of this field.

Luís Cruz, Rui Abreu

High energy consumption is a challenging issue that an ever increasing number of mobile applications face today. However, energy consumption is being tested in an ad hoc way, despite being an important non-functional requirement of an application. Such limitation becomes particularly disconcerting during software testing: on the one hand, developers do not really know how to measure energy; on the other hand, there is no knowledge as to what is the energy overhead imposed by the testing framework. In this paper, as we evaluate eight popular mobile UI automation frameworks, we have discovered that there are automation frameworks that increase energy consumption up to roughly 2200%. While limited in the interactions one can do, Espresso is the most energy efficient framework. However, depending on the needs of the tester, Appium, Monkeyrunner, or UIAutomator are good alternatives. In practice, results show that deciding which is the most suitable framework is vital. We provide a decision tree to help developers make an educated decision on which framework suits best their testing needs.