Software Changes Research Articles

Generative AI for software testing: Harnessing large language models for automated and intelligent quality assurance

Software testing is indispensable for ensuring that modern applications meet rigorous standards of functionality, reliability, and security. However, the complexity and pace of contemporary software development often overwhelm traditional and even AI-based testing approaches, leading to gaps in coverage, delayed feedback, and increased maintenance costs. Recent breakthroughs in Generative AI, particularly Large Language Models (LLMs), offer a new avenue for automating and optimizing testing processes. These models can dynamically generate test cases, predict system vulnerabilities, handle continuous software changes, and reduce the burden on human testers. This paper explores how Generative AI complements and advances established AI-driven testing frameworks, outlines the associated challenges of data preparation and governance, and proposes future directions for fully autonomous, trustworthy testing solutions.

Assessing the stability of photon-counting CT: insights from a 2-year longitudinal study.

Among the advancements in computed tomography (CT) technology, photon-counting computed tomography (PCCT) stands out as a significant innovation, providing superior spectral imaging capabilities while simultaneously reducing radiation exposure. Its long-term stability is important for clinical care, especially longitudinal studies, but is currently unknown. This study sets out to comprehensively analyze the long-term stability of a first-generation clinical PCCT scanner. Over a 2-year period, from November 2021 to November 2023, we conducted weekly identical experiments utilizing the same multi-energy CT protocol. Throughout this period, notable software and hardware modifications were meticulously recorded. Various tissue-mimicking inserts were scanned weekly to rigorously assess the stability of Hounsfield Units (HU) and image noise in Virtual Monochromatic Images (VMIs) and iodine density maps. Spectral results consistently demonstrated the quantitative stability of PCCT. VMIs exhibited stable HU values, such as variation in relative error for VMI 70 keV measuring 0.11% and 0.30% for single-source and dual-source modes, respectively. Similarly, noise levels remained stable with slight fluctuations linked to software changes for VMI 40 and 70 keV that corresponded to changes of 8 and 1 HU, respectively. Furthermore, iodine density quantification maintained stability and showed significant improvement with software and hardware changes, especially in dual-source mode with nominal errors decreasing from 1.44 to 0.03 mg/mL. This study provides the first long-term reproducibility assessment of quantitative PCCT imaging, highlighting its potential for the clinical arena. Question Photon-counting CT (PCCT) provides critical spectral imaging for improved diagnostic accuracy, but its long-term quantitative stability over time is still unknown. Findings The clinical PCCT system demonstrated stable Hounsfield Units (HU) and image noise over 2 years, ensuring reliable quantitative imaging and improving diagnostic accuracy. Clinical relevance This study showcased the exceptional value of PCCT in diagnostic radiology, particularly for its application in longitudinal studies.

Research on Methods for Developing Software Identification Codes in Automotive Software Update Management Regulations

As vehicles become increasingly digitized and intelligent, with growing demands for functional iterations, software updates (OTA) are becoming more frequent. National regulatory agencies have taken corresponding measures to address the risks associated with automotive software updates. To enhance the automotive industry's standards and maintaina healthy industry ecosystem, it is urgent to establish requirements for managing software identification codes and tracking software changes. Current software update management documents, primarily the UN R156 regulation and the forthcoming GB "General Technical Requirements for Automotive Software Updates," specify requirements for software identification codes. Based on the UN R156 regulation and research on existing automotive code systems in China, we propose ideas and methods for developing software identification codes.

Experimental verification of changes in the control map of a diesel engine operation due to fuel consumption

The paper describes an experiment related to the introduction of changes in the control software of a supercharged diesel engine in order to reduce fuel consumption, as a measure of useful efficiency. The map shaping the torque in the range of low and medium engine speeds was modified. The fuel dose for a specific engine speed at maximum engine load was changed and the boost pressure was increased, without exceeding the permissible values. A vehicle from the age group of 16 years - typical for the automotive market in Poland, was selected for the tests. It meets the EURO 5 standard. The experiment was carried out on a chassis dynamometer and in road traffic conditions. The average fuel consumption was obtained lower by 0.4 dm3 per 100 km than the factory assumed.

Carving Out Control Code: Automated Identification of Control Software in Autopilot Systems

Cyber-physical systems interact with the world through software controlling physical effectors. Carefully designed controllers, implemented as safety-critical control software, also interact with other parts of the software suite, and may be difficult to separate, verify, or maintain. Moreover, some software changes, not intended to impact control system performance, do change controller response through a variety of means including interaction with external libraries or unmodeled changes only existing in the cyber system (e.g., exception handling). As a result, identifying safety-critical control software, its boundaries with other embedded software in the system, and the way in which control software evolves could help developers isolate, test, and verify control implementation, and improve control software development. In this work we present an automated technique, based on a novel application of machine learning, to detect commits related to control software, its changes, and how the control software evolves. We leverage messages from developers (e.g., commit comments), and code changes themselves to understand how control software is refined, extended, and adapted over time. We examine three distinct, popular, real-world, safety-critical autopilots—ArduPilot, Paparazzi UAV, and LibrePilot to test our method demonstrating an effective detection rate of 0.95 for control-related code changes.

Software Change Request in Software Development Project : Factors and Methods (Scoping Review Methods)

Software development is dynamic by nature, often requiring changes in software requirements through Software Change Requests (SCRs). This research aims to identify and analyze the key factors that influence SCRs in software development projects and evaluate the methods used to effectively manage these changes. This research uses the Scoping Review Method to systematically review and synthesize existing literature on SCRs, aiming to provide a comprehensive understanding of current practices, challenges, and strategies for managing software change requests. The review includes an extensive search of academic databases to identify relevant studies that address SCR factors and management methods. Key factors that influence SCR include project size, complexity, stakeholder involvement, and clarity of initial requirements. Methods for managing SCR are categorized into formal approaches, such as structured change control processes and tools, and informal approaches, which rely on ad-hoc communication and collaboration. The results underscore the importance of a well-defined change management process to mitigate the risks associated with SCR. Effective SCR management can improve project outcomes by improving software quality, increasing stakeholder satisfaction, and minimizing project delays. This research has practical implications for software development professionals by providing insight into effective SCR management techniques. This research also identifies gaps in the literature, recommending future research on the impact of agile methodologies and the use of automated tools in facilitating change management. Understanding these factors and methods enables practitioners to better navigate the complexities of software change, ensuring successful project delivery and maintaining high software quality.

A Meta-Study of Software-Change Intentions

Every software system undergoes changes, for example, to add new features, fix bugs, or refactor code. The importance of understanding software changes has been widely recognized, resulting in various techniques and studies, for instance, on change-impact analysis or classifying developers’ activities. Since changes are triggered by developers’ intentions—something they plan or want to change in the system—many researchers have studied intentions behind changes. While there appears to be a consensus among software-engineering researchers and practitioners that knowing the intentions behind software changes is important, it is not clear how developers can actually benefit from this knowledge. In fact, there is no consolidated, recent overview of the state of the art on software-change intentions (SCIs) and their relevance for software engineering. We present a meta-study of 122 publications, which we used to derive a categorization of SCIs and to discuss motivations, evidence, and techniques relating to SCIs. Unfortunately, we found that individual pieces of research are often disconnected from each other, because a common understanding is missing. Similarly, some publications showcase the potential of knowing SCIs, but more substantial research to understand the practical benefits of knowing SCIs is needed. Our contributions can help researchers and practitioners improve their understanding of SCIs and how SCIs can aid software engineering tasks.

Mechatronic technological system based on CNC and intelligent mechatronic mechanism

A mechatronic technological system (MTS) has been developed. It consisting of a CNC metal-cutting machine tool and a mechatronic spindle, made on the basis of an intelligent mechatronic mechanism (IMM). The main elements of the technological system – the CNC and IMM devices – are parts of a single hierarchical, two-level control system, representing, respectively, the upper and lower control levels in this system. The upper control level – the CNC device – ensures stabilization of the kinematic parameters of machining (displacement, speed, acceleration) and their software change in accordance with the CNC control program. The lower control level – the IMM device – ensures stabilization of the power parameters of mechanical machining (cutting axial force and torque) and their software change in accordance with the technological requirements for the surface quality and the physical-and-mechanical state of the surface layer of the machined parts. At both control levels (the upper and lower ones), closed automatic control systems “by deviation” of the adjustable kinematic (upper control level) and power (lower control level) parameters of mechanical machining are functioning. Moreover, the automatic control system of the lower control level of the hierarchy in relation to a similar automatic control system of the upper control level, implements the method of automatic control “by disturbance”. This, in addition to directly ensuring surface finish and surface integrity, allows improving the quality of stabilization of kinematic cutting parameters at the upper control level. Experimental studies of MTS were carried out when machining parts from various materials (steel, polymer composite materials, superhard crystals and stones, other difficult-to-machine materials).

A case report of system configuration issue in medical imaging due to system upgrade- changes in hardware and software.

Although the rapid growth in the efficiency of medical imaging is undeniable, the expansion of health information technology (HIT) into medical imaging has not been as seamless or well-integrated as it was thought to be. The socio-technical complexities in medical imaging associated with HIT systems can cause risks to patient harm and inconvenience, both individually and collectively, often in new, unforeseen, and unexpected ways. This study reflects a retrospectively collected single incident report related to medical imaging HIT systems, aiming to develop a set of preventive and corrective strategies. A combination of multiple deductive approaches (existing frameworks), i.e., HIT Classification Systems and 18-step medical imaging process workflow and inductive method (content analysis), were used to analyze the incident. The incident was identified as a "system configuration"-related software issue, contributed by system upgrade- changes in hardware and software. The incident was determined to occur during steps 10-12, i.e., "study selection and retrieval," "calling up of patient's referral," and "image review and interpretation," causing severe disruptions in the clinical workflow for several weeks. We propose 16 preventive and corrective strategies grouped under four key areas based on the socio-technical aspects associated with HIT systems. The key areas are (i) preparation and integration for upgraded systems, (ii) training for medical imaging specialists, (iii) contingency planning/immediate backup system, and (iv) system design and configuration. These strategies are expected to help healthcare staff, analysts, reporters, researchers, and relevant stakeholders improve care delivery and patient safety in medical imaging in the context of any system upgrades.

ChangeRCA: Finding Root Causes from Software Changes in Large Online Systems

ChangeRCA: Finding Root Causes from Software Changes in Large Online Systems

Assessing the Stability of Photon-Counting CT: Insights from a Two-Year Longitudinal Study.

Among the advancements in computed tomography (CT) technology, photon-counting computed tomography (PCCT) stands out as a significant innovation, providing superior spectral imaging capabilities while simultaneously reducing radiation exposure. Its long-term stability is important for clinical care, especially longitudinal studies, but is currently unknown. This study sets out to comprehensively analyze the long-term stability of a first-generation clinical PCCT scanner. Over a two-year period, from November 2021 to November 2023, we conducted weekly identical experiments utilizing the same multi-energy CT protocol. These experiments included various tissue-mimicking inserts to rigorously assess the stability of Hounsfield Units (HU) and image noise in Virtual Monochromatic Images (VMIs) and iodine density maps. Throughout this period, notable software and hardware modifications were meticulously recorded. Each week, VMIs and iodine density maps were reconstructed and analyzed to evaluate quantitative stability over time. Spectral results consistently demonstrated the quantitative stability of PCCT. VMIs exhibited stable HU values, such as variation in relative error for VMI 70 keV measuring 0.11% and 0.30% for single-source and dual-source modes, respectively. Similarly, noise levels remained stable with slight fluctuations linked to software changes for VMI 40 and 70 keV that corresponded to changes of 8 and 1 HU, respectively. Furthermore, iodine density quantification maintained stability and showed significant improvement with software and hardware changes, especially in dual-source mode with nominal errors decreasing from 1.44 to 0.03 mg/mL. This study provides the first long-term reproducibility assessment of quantitative PCCT imaging, highlighting its potential for the clinical arena. This study indicates its long-term utility in diagnostic radiology, especially for longitudinal studies.

Suppression of Subpixel Jitter in Resonant Scanning Systems With Phase-locked Sampling.

Resonant scanning is critical to high speed and in vivo imaging in many applications of laser scanning microscopy. However, resonant scanning suffers from well-known image artifacts due to scanner jitter, limiting adoption of high-speed imaging technologies. Here, we introduce a real-time, inexpensive and all electrical method to suppress jitter more than an order of magnitude below the diffraction limit that can be applied to most existing microscope systems with no software changes. By phase-locking imaging to the resonant scanner period, we demonstrate an 86% reduction in pixel jitter, a 15% improvement in point spread function with resonant scanning and show that this approach enables two widely used models of resonant scanners to achieve comparable accuracy to galvanometer scanners running two orders of magnitude slower. Finally, we demonstrate the versatility of this method by retrofitting a commercial two photon microscope and show that this approach enables significant quantitative and qualitative improvements in biological imaging.

Interaction of integral parameters of physical state and biological age of students aged 17-19 years old in the process of adaptation to standard and experimental physical education programs

Background and Study Aim. The process of adaptation to physical education programs is a critical factor influencing the health and physical development of students. Understanding how these programs affect both the physical state and biological age of students can significantly improve their effectiveness. The aim of the study is to investigate the interaction between integral parameters of physical state (PS) and biological age (BA) of students aged 17-19 years old in the process of adaptation to standard and experimental physical education (PE) programs. Material and Methods. The study involved 140 first-year students of the National University of Telecommunications - 80 males and 60 females. In the course of the research, the indices of functional state, physical fitness (PF) and physical development (PD), biological age (BA) were recorded in young men and women of the control (CG) and experimental (EG) groups. To process the experimental material, methods of dispersion and canonical analysis were used and multiple correlation and determination coefficients were calculated. Results. It was revealed that the impact of PS integral parameters on students' BA is stronger than the inverse influence of BA on PS parameters. The degree of BA reduction under the influence of the studied factors is most pronounced in the EG of girls and boys. The variability of the group of PD and PF indices is accompanied by the variability of most of the BA parameters’ dispersion in all studied groups of students. The factor of gender has the greatest impact on the BA of students. Girls have lower initial values of BA and it’s a more pronounced decrease by the end of the experiment. The effectiveness of the impact of an aerobically oriented experimental program on the improvement of students’ PS and reduction of BA was confirmed. Under the influence of the experimental program, the BA of EG girls decreases to a greater extent in comparison with EG boys and students of both control groups. Changes and differences in BA of boys and girls of EG and CG are due to the combined impact of the factor of PE software and that of adaptation changes of students' PS in the annual process of PE. In the group of boys and girls more pronounced influence on BA in the course of the experiment was produced by the factor of software and adaptation changes, respectively. Conclusions. The gender factor has the strongest impact on students' BA - girls have lower initial values of BA and it’s a more pronounced decrease by the end of the experiment in comparison with boys. Adaptation changes of PS in the process of software implementation are accompanied by a decrease in BA in the experimental groups. One of the criteria and confirmations of the effectiveness of the developed PE program is the degree of mutual influence of PS integral parameters, which is most pronounced in the reduction of BA in the experimental groups of girls and boys.

Estimating Uncertainty in Labeled Changes by SZZ Tools on Just-In-Time Defect Prediction

The aim of Just-In-Time (JIT) defect prediction is to predict software changes that are prone to defects in a project in a timely manner, thereby improving the efficiency of software development and ensuring software quality. Identifying changes that introduce bugs is a critical task in just-in-time defect prediction, and researchers have introduced the SZZ approach and its variants to label these changes. However, it has been shown that different SZZ algorithms introduce noise to the dataset to a certain extent, which may reduce the predictive performance of the model. To address this limitation, we propose the Confident Learning Imbalance (CLI) model. The model identifies and excludes samples whose labels may be corrupted by estimating the joint distribution of noisy labels and true labels, and mitigates the impact of noisy data on the performance of the prediction model. The CLI consists of two components: identifying noisy data (Confident Learning Component) and generating a predicted probability matrix for imbalanced data (Imbalanced Data Probabilistic Prediction Component). The IDPP component generates precise predicted probabilities for each instance in the training set, while the CL component uses the generated predicted probability matrix and noise labels to clean up the noise and build a classification model. We evaluate the performance of our model through extensive experiments on a total of 126,526 changes from ten Apache open source projects, and the results show that our model outperforms the baseline methods.

Performance changes due to differences among annotating radiologists for training data in computerized lesion detection

PurposeThe quality and bias of annotations by annotators (e.g., radiologists) affect the performance changes in computer-aided detection (CAD) software using machine learning. We hypothesized that the difference in the years of experience in image interpretation among radiologists contributes to annotation variability. In this study, we focused on how the performance of CAD software changes with retraining by incorporating cases annotated by radiologists with varying experience.MethodsWe used two types of CAD software for lung nodule detection in chest computed tomography images and cerebral aneurysm detection in magnetic resonance angiography images. Twelve radiologists with different years of experience independently annotated the lesions, and the performance changes were investigated by repeating the retraining of the CAD software twice, with the addition of cases annotated by each radiologist. Additionally, we investigated the effects of retraining using integrated annotations from multiple radiologists.ResultsThe performance of the CAD software after retraining differed among annotating radiologists. In some cases, the performance was degraded compared to that of the initial software. Retraining using integrated annotations showed different performance trends depending on the target CAD software, notably in cerebral aneurysm detection, where the performance decreased compared to using annotations from a single radiologist.ConclusionsAlthough the performance of the CAD software after retraining varied among the annotating radiologists, no direct correlation with their experience was found. The performance trends differed according to the type of CAD software used when integrated annotations from multiple radiologists were used.

Methods for analyzing the impact of software changes on objective functions and safety functions

Objectives. This paper examines the various approaches to analyzing the impact of software changes, and suggests a new method using function control flows. Impact analysis of software change can require the investment of a lot of time and competence on the part of the expert conducting it. There is no detailed description of methodology for analyzing the impact of changes and it is not established at a legislative level. The proposed method has three aims: reducing the level of requirements for an expert when conducting software research; localizing code areas to establish defects in information protection functions; and reducing the time spent on analyzing the impact of changes.Methods. The study analyzes the common methods for analyzing software changes with a description of their positive and negative sides. The possibility of analyzing changes in the control flow of software functions is considered as an alternative to line-by-line comparison of the full volume of source codes. Represented as tree-shaped graphs, the control flows of different versions of the same software are subject to a merging procedure. The final result is analyzed by an expert from the research organization.Results. The research results of the software change analysis methods are presented with a description of their disadvantages. A description is given of the method for change analysis using function control. This complements existing methods, while eliminating their disadvantages. The study also analyzes the possibility of using this method beyond the tasks defined in the introduction.Conclusions. The use of methods to localize the most vulnerable code sections is considered one of the most promising areas for analyzing change impact. In addition to searching for vulnerable code sections, it is important to evaluate the effectiveness of the control flow comparison method in the analysis of source code when transferred to another code base.

Comparing Single-Tooth Implant Placement With and Without Allogeneic Soft Tissue Augmentation: A Randomized Clinical Trial.

To explore the effect of adding an allogeneic soft tissue graft at the time of single implant placement using a fully digital workflow for implant placement and restoration without making either analog or digital impressions. A prospective randomized clinical study was performed enrolling 39 participants requiring single-tooth implants. The patients were randomized into one of two groups: (1) the graft group, which received an allogeneic dermal graft at the time of implant placement (n = 19), or (2) the nongraft group (n = 20). A fully digital surgical and restorative protocol was implemented for both groups. Intraoral scans were taken before implant placement (T0), at the time of final crown delivery (T1), and 1 year after placement (T2). Intraoral scans were aligned using Geomagic Control X 2020 software, and linear and volumetric changes in buccal tissues were measured at T0, T1, and T2. Implant survival, probing depths (PDs), and complications were recorded. Participants were asked to complete an Oral Health Impact Profile (OHIP)- 14 survey at T0 and T2. Marginal bone levels were measured at T0 and T2 on periapical radiographs. Overall, 39 participants completed surgery and restoration in the incisor, canine, premolar, and molar sites. Two early failures were recorded in central incisor positions (95% survival). Crown delivery without complication from the digital workflow (impressionless) was achieved for 36 out of 39 cases (92%), and implant depth control was the main challenge. At 1 year, 37 participants attended the follow-up appointment. Both groups showed gain in buccal tissue thickness without significant differences between the two groups for both linear and volumetric measurements (P > .05). Soft tissue grafting was associated with minimal added morbidity. The interproximal marginal bone changes were recorded as follows: -0.16 mm mesial and -0.12 mm distal for the graft group and -0.01 mm mesial and -0.11 mm distal for the nongraft group (P = .07 for mesial and .83 for distal). The OHIP score was significantly reduced at T2 compared to T0 (P = .003) for the entire cohort. Augmentation of the alveolar mucosa on the buccal aspect of single-tooth implants is associated with clinically favorable outcomes. A fully digital workflow has been validated to permit crown delivery on CAD/CAM abutments without implant impressions.

7 Dimensions of software change patterns

Evolving software is a highly complex and creative problem in which a number of different strategies are used to solve the tasks at hand. These strategies and reoccurring coding patterns can offer insights into the process. However, they can be highly project or even task-specific. We aim to identify code change patterns in order to draw conclusions about the software development process. For this, we propose a novel way to calculate high-level file overarching diffs, and a novel way to parallelize pattern mining. In a study of 1000 Java projects, we mined and analyzed a total of 45,000 patterns. We present 13 patterns, showing extreme points of the 7 pattern categories we identified. We found that a large number of high-level change patterns exist and occur frequently. The majority of mined patterns were associated with a specific project and contributor, where and by whom it was more likely to be used. While a large number of different code change patterns are used, only a few, mostly unsurprising ones, are common under all circumstances. The majority of code change patterns are highly specific to different context factors that we further explore.

Influences of Software Changes on Oxycodone Prescribing at an Australian Tertiary Emergency Department: A Retrospective Review.

Opioid prescribing and dispensing from emergency departments is a noteworthy issue given widespread opioid misuse and diversion in many countries, contributing both physical and economic harm to the population. High patient numbers and the stochastic nature of acute emergency presentations to emergency departments (EDs) introduce challenges for prescribers who are considering opioid stewardship principles. This study investigated the effect of changes to electronic prescribing software on prescriptions with an auto-populated quantity of oxycodone immediate release (IR) from an Australian tertiary emergency department following the implementation of national recommendations for reduced pack sizes. A retrospective review of oxycodone IR prescriptions over two six-month periods between 2019 and 2021 was undertaken, either side of a software adjustment to reduce the default quantities of tablets prescribed from 20 to 10. Patient demographic details were collected, and prescriber years of practice calculated for inclusion in linear mixed effects regression modelling. A reduction in the median number of tablets prescribed per prescription following the software changes (13.5 to 10.0, p < 0.001) with little change in the underlying characteristics of the patient or prescriber populations was observed, as well as an 11.65% reduction in the total number of tablets prescribed. The prescriber's years of practice, patient age and patient sex were found to influence increased prescription sizes. Reduced quantity of oxycodone tablets prescribed was achieved by alteration of prescribing software prefill parameters, providing further evidence to support systems-based policy interventions to influence health care providers behaviour and to act as a forcing function for prescribers to consider opioid stewardship principles.

An assessment of heterogenous ensemble classifiers for analyzing change‐proneness in open‐source software systems

AbstractSoftware managers constantly look out for methods that ensure cost effective development of good quality software products. An important means of accomplishing this is by allocating more resources to weak classes of a software product, which are prone to changes. Therefore, correct prediction of these change‐prone classes is critical. Though various researchers have investigated the performance of several algorithms for identifying them, the search for an optimum classifier still persists. To this end, this study critically investigates the use of six Heterogenous Ensemble Classifiers (HEC) for Software Change Prediction (SCP) by empirically validating datasets obtained from 12 open‐source software systems. The results of the study are statistically assessed using three robust performance indicators (AUC, F‐measure and Mathew Correlation Coefficient) in two different validation scenarios (within project and cross‐project). They indicate the superiority of Average Probability Voting Ensemble, a heterogenous classifier for determining change‐proneness in the investigated systems. The average AUC values of software change prediction models developed using this ensemble classifier exhibited an improvement of 3%‐9% and 3%‐11% respectively when compared with its base learners and homogeneous counter parts. Similar observations were inferred using other investigated performance measures. Furthermore, the evidence obtained from the results suggests that the change in number of base learners or type of meta‐learner does not exhibit significant change in the performance of corresponding heterogenous ensemble classifiers.