https://www.mdu.se/

Recommender systems for requirements are typically built on the assumption that similar requirements can be used as proxies to retrieve similar software. When a stakeholder proposes a new requirement, natural language processing (NLP)-based similarity metrics can be exploited to retrieve existing requirements, and in turn, identify previously developed code. Several NLP approaches for similarity computation between requirements are available. However, there is little empirical evidence on their effectiveness for code retrieval. This study compares different NLP approaches, from lexical ones to semantic, deep-learning techniques, and correlates the similarity among requirements with the similarity of their associated software. The evaluation is conducted on real-world requirements from two industrial projects from a railway company. Specifically, the most similar pairs of requirements across two industrial projects are automatically identified using six language models. Then, the trace links between requirements and software are used to identify the software pairs associated with each requirements pair. The software similarity between pairs is then automatically computed with JPLag. Finally, the correlation between requirements similarity and software similarity is evaluated to see which language model shows the highest correlation and is thus more appropriate for code retrieval. In addition, we perform a focus group with members of the company to collect qualitative data. Results show a moderately positive correlation between requirements similarity and software similarity, with the pre-trained deep learning-based BERT language model with preprocessing outperforming the other models. Practitioners confirm that requirements similarity is generally regarded as a proxy for software similarity. However, they also highlight that additional aspect comes into play when deciding software reuse, e.g., domain/project knowledge, information coming from test cases, and trace links. Our work is among the first ones to explore the relationship between requirements and software similarity from a quantitative and qualitative standpoint. This can be useful not only in recommender systems but also in other requirements engineering tasks in which similarity computation is relevant, such as tracing and change impact analysis.

Requirements prioritization plays an important role in driving project success during software development. Literature reveals that existing requirements prioritization approaches ignore vital factors such as interdependency between requirements. Existing requirements prioritization approaches are also generally time-consuming and involve substantial manual effort. Besides, these approaches show substantial limitations in terms of the number of requirements under consideration. There is some evidence suggesting that models could have a useful role in the analysis of requirements interdependency and their visualization, contributing towards the improvement of the overall requirements prioritization process. However, to date, just a handful of studies are focused on model-based strategies for requirements prioritization, considering only conflict-free functional requirements. This paper uses a meta-model-based approach to help the requirements analyst to model the requirements, stakeholders, and inter-dependencies between requirements. The model instance is then processed by our modified PageRank algorithm to prioritize the given requirements. An experiment was conducted, comparing our modified PageRank algorithm’s efficiency and accuracy with five existing requirements prioritization methods. Besides, we also compared our results with a baseline prioritized list of 104 requirements prepared by 28 graduate students. Our results show that our modified PageRank algorithm was able to prioritize the requirements more effectively and efficiently than the other prioritization methods.

Over the years, a range of Internet of Vehicles services has emerged, along with improved quality parameters. However, the field still faces several limitations, including resource constraints and the time response requirement. This paper extracts cost, energy, processing power, service management, and resource allocation parameters. Mathematical equations are then defined based on these parameters. To simplify the process complexity and ensure scalability, we propose an algorithm that uses the genetic algorithm for fault and cost management during resource allocation to services. The main concept is to pick resources for services using a genetic algorithm. We discuss the processing and energy costs associated with this function, which is the algorithm's objective function and is created to optimize cost. Our approach goes beyond the conventional genetic algorithm in two stages. In the first step, services are prioritized, and resources are allocated in accordance with those priorities; in the second step, load balancing in message transmission paths is ensured, and message failures are avoided. The algorithm's performance is evaluated using various parameters, and it was shown to outperform other metaheuristic algorithms like the classic genetic algorithm, particle swarm, and mathematical models. Different scenarios with various nodes and service variables are defined in various system states, including fault occurrences to various percentages of 10, 20, and 30. To compare methods, we consider different parameters, the most significant being performance success rate. Moreover, the cost optimization has a good convergence after iterations, and the rate of improvement in the big scenario has slowed down after 150 iterations. Besides, it provides acceptable performance in response time for services.

Industrial cybersecurity has risen as an important topic of research nowadays. The heavy connectivity by the Internet of Things (IoT) and the growth of cyberattacks against industrial assets cause this risen and attract attention to the cybersecurity field. While fostering current software applications and use-cases, the ubiquitous access to the Internet has also exposed operational technologies to new and challenging security threats that need to be addressed. As the number of attacks increases, their visibility decreases. An attack can modify the Cyber-Physical Systems (CPSs) quality to avoid proper quality assessment. They can disrupt the system design process and adversely affect a product’s design purpose. This working progress paper presents our approach to modeling, analyzing, and mitigating cyberattacks in CPS. We model the normal behavior of the application as well as cyberattacks with the help of Microsoft Security Development Lifecycle (SDL) and threat modeling approach (STRIDE). Then verify the application and attacks model using a model checking tool and propose mitigation strategies to decrease the risk of vulnerabilities. The results can be used to improve the system design to overcome the vulnerabilities.

This article presents a runtime verification tool for embedded software executing under the open source real-time operating system FreeRTOS. The tool detects and diagnoses concurrency bugs such as deadlock, starvation, and suspension based-locking. The tool finds concurrency bugs at runtime without debugging and tracing the source code. The tool uses the Tracealyzer tool for logging relevant events. Analysing the logs, our tool can detect the concurrency bugs by applying algorithms for diagnosing each concurrency bug type individually. In this paper, we present the implementation of the tool, as well as its functional architecture, together with illustration of its use. The tool can be used during program testing to gain interesting information about embedded software executions. We present initial results of running the tool on some classical bug examples running on an AVR 32-bit board SAM4S. 

Non-invasive monitoring of cardiovascular health through photoplethysmography (PPG) waveforms has emerged as a crucial area of research. The Windkessel 4-Element (WK4) model is a mathematical approach used to estimate key physiological parameters related to cardiovascular health, including arterial compliance, peripheral resistance, inertance, and total arterial resistance. This study aimed to evaluate key physiological parameters associated with cardiovascular health using the WK4 model, leveraging real-life PPG waveform data obtained from volunteers across three distinct age groups. To achieve this, an algorithm was developed to automatically determine optimal parameter values for each volunteer. The results revealed a mean correlation coefficient of 0.96 between the automatically generated waveforms by the algorithm and the actual real-life PPG waveforms, indicating robust agreement. Notably, only the total arterial resistance parameter exhibited significant differences among the age groups, suggesting that the algorithm holds promise for detecting agerelated changes in cardiovascular health. These findings emphasize the potential for the development of a non-invasive tool to assess cardiovascular health status and enhance healthcare outcomes. Furthermore, they underscore the capability of the developed algorithm as a non-invasive means to evaluate various aspects of cardiovascular physiology. Additionally, the versatility of this algorithm opens doors for its application in educational settings, promoting knowledge advancement, empowering research endeavors, and facilitating advancements in the field.

Accurate bladder monitoring is critical in the management of conditions such as urinary incontinence, voiding dysfunction, and spinal cord injuries. Electrical bioimpedance (EBI) has emerged as a cost-effective and non-invasive approach to monitoring bladder activity in daily life, with particular relevance to patient groups who require measurement of bladder urine volume (BUV) to prevent urinary leakage. However, the impact of activities in daily living (ADLs) on EBI measurements remains incompletely characterized. In this study, we investigated the impact of normal ADLs such as sitting, standing, and walking on EBI measurements using the MAX30009evkit system with four electrodes placed on the lower abdominal area. We developed an algorithm to identify artifacts caused by the different activities from the EBI signals. Our findings demonstrate that various physical activities clearly affected the EBI measurements, indicating the necessity of considering them during bladder monitoring with EBI technology performed during physical activity (or normal ADLs). We also observed that several specific activities could be distinguished based on their impedance values and waveform shapes. Thus, our results provide a better understanding of the impact of physical activity on EBI measurements and highlight the importance of considering such physical activities during EBI measurements in order to enhance the reliability and effectiveness of EBI technology for bladder monitoring.

Bladder monitoring, including urinary incontinence management and bladder urinary volume monitoring, is a vital part of urological care. Urinary incontinence is a common medical condition affecting the quality of life of more than 420 million people worldwide, and bladder urinary volume is an important indicator to evaluate the function and health of the bladder. Previous studies on non-invasive techniques for urinary incontinence management technology, bladder activity and bladder urine volume monitoring have been conducted. This scoping review outlines the prevalence of bladder monitoring with a focus on recent developments in smart incontinence care wearable devices and the latest technologies for non-invasive bladder urine volume monitoring using ultrasound, optical and electrical bioimpedance techniques. The results found are promising and their application will improve the well-being of the population suffering from neurogenic dysfunction of the bladder and the management of urinary incontinence. The latest research advances in bladder urinary volume monitoring and urinary incontinence management have significantly improved existing market products and solutions and will enable the development of more effective future solutions.

Development in the field of smart wearable products for monitoring daily life health status is beginning to spread in society. Textile electronic methods are improving and facilitating the manufacturing of sensorized garments. This paper evaluates a newly developed t-shirt incorporating electronic sensing and interconnecting elements integrated into the T-shirt with textile-friendly techniques sensorized with a Movesense device for monitoring ECG and HR and activity. The measurement results obtained from the t-shirt are entirely in agreement with the measurements obtained with other textile garments and encourage us for a near future where wearable sensors are just textile garments sensorized seamlessly without suboptimal textile-electronic integrated elements.

The development of smart wearable solutions for monitoring daily life health status is increasingly popular, with chest straps and wristbands being predominant. This study introduces a novel sensorized T-shirt design with textile electrodes connected via a knitting technique to a Movesense device. We aimed to investigate the impact of stationary and movement actions on electrocardiography (ECG) and heart rate (HR) measurements using our sensorized T-shirt. Various activities of daily living (ADLs), including sitting, standing, walking, and mopping, were evaluated by comparing our T-shirt with a commercial chest strap. Our findings demonstrate measurement equivalence across ADLs, regardless of the sensing approach. By comparing ECG and HR measurements, we gained valuable insights into the influence of physical activity on sensorized T-shirt development for monitoring. Notably, the ECG signals exhibited remarkable similarity between our sensorized T-shirt and the chest strap, with closely aligned HR distributions during both stationary and movement actions. The average mean absolute percentage error was below 3%, affirming the agreement between the two solutions. These findings underscore the robustness and accuracy of our sensorized T-shirt in monitoring ECG and HR during diverse ADLs, emphasizing the significance of considering physical activity in cardiovascular monitoring research and the development of personal health applications.

In this paper, we present a hierarchical model for time predictable task scheduling in edge-cloud computing architecture for industrial cyber-physical systems. Regarding the scheduling problem, we also investigate the common problem-solving approaches and discuss our preliminary plan to realize the proposed architecture. Furthermore, an Integer linear programming (ILP) model is proposed for task scheduling problem in the cloud layer. The model considers timing and security requirements of applications and the objective is to minimize the financial cost of their execution.

The edge-cloud computing continuum effectively uses fog and cloud servers to meet the quality of service (QoS) requirements of tasks when edge devices cannot meet those requirements. This paper focuses on the workflow offloading problem in edge-cloud computing and formulates this problem as a nonlinear mathematical programming model. The objective function is to minimize the monetary cost of executing a workflow while satisfying constraints related to data dependency among tasks and QoS requirements, including security and deadlines. Additionally, it presents a genetic algorithm for the workflow offloading problem to find near-optimal solutions with the cost minimization objective. The performance of the proposed mathematical model and genetic algorithm is evaluated on several real-world workflows. Experimental results demonstrate that the proposed genetic algorithm can find admissible solutions comparable to the mathematical model and outperforms particle swarm optimization, bee life algorithm, and a hybrid heuristic-genetic algorithm in terms of workflow execution costs.

A hybrid cloud is an efficient solution to deal with the problem of insufficient resources of a private cloud when computing demands increase beyond its resource capacities. Cost-efficient workflow scheduling, considering security requirements and data dependency among tasks, is a prominent issue in the hybrid cloud. To address this problem, we propose a mathematical model that minimizes the monetary cost of executing a workflow and satisfies the security requirements of tasks under a deadline. The proposed model fulfills data dependency among tasks, and data transmission time is formulated with exact mathematical expressions. The derived model is a Mixed-integer linear programming problem. We evaluate the proposed model with real-world workflows over changes in the input variables of the model, such as the deadline and security requirements. This paper also presents a post-optimality analysis that investigates the stability of the assignment problem. The experimental results show that the proposed model minimizes the cost by decreasing inter-cloud communications for dependent tasks. However, the optimal solutions are affected by the limitations that are imposed by the problem constraints. 

Task offloading is a prominent problem in edge−cloud computing, as it aims to utilize the limited capacityof fog servers and cloud resources to satisfy the QoS requirements of tasks, such as meeting their deadlines.This paper formulates the task offloading problem as a nonlinear mathematical programming model to maximizethe number of independent IoT tasks that meet their deadlines and to minimize the deadline violationtime of tasks that cannot meet their deadlines. This paper proposes two Q-learning algorithms to solve theformulated problem. The performance of the proposed algorithms is experimentally evaluated with respect toseveral algorithms. The evaluation results demonstrate that the proposed Q-learning algorithms perform wellin meeting task deadlines and reducing the total deadline violation time.

Cardiovascular diseases (CVDs) are a leading cause of death worldwide, and hypertension is a major risk factor for acquiring CVDs. Early detection and treatment of hypertension can significantly reduce the risk of developing CVDs and related complications. In this study, a linear SVM machine learning model was used to classify subjects as normal or at different stages of hypertension. The features combined statistical parameters derived from the acceleration plethysmography waveforms and clinical parameters extracted from a publicly available dataset. The model achieved an overall accuracy of 87.50% on the validation dataset and 95.35% on the test dataset. The model's true positive rate and positive predictivity was high in all classes, indicating a high accuracy, and precision. This study represents the first attempt to classify cardiovascular conditions using a combination of acceleration photoplethysmogram (APG) features and clinical parameters The study demonstrates the potential of APG analysis as a valuable tool for early detection of hypertension.

The extraction of relevant features from the photoplethysmography signal for estimating certain physiological parameters is a challenging task. Various feature extraction methods have been proposed in the literature. In this study, we present a novel fiducial point extraction algorithm to detect c and d points from the acceleration photoplethysmogram (APG), namely “CnD”. The algorithm allows for the application of various pre-processing techniques, such as filtering, smoothing, and removing baseline drift; the possibility of calculating first, second, and third photoplethysmography derivatives; and the implementation of algorithms for detecting and highlighting APG fiducial points. An evaluation of the CnD indicated a high level of accuracy in the algorithm’s ability to identify fiducial points. Out of 438 APG fiducial c and d points, the algorithm accurately identified 434 points, resulting in an accuracy rate of 99%. This level of accuracy was consistent across all the test cases, with low error rates. These findings indicate that the algorithm has a high potential for use in practical applications as a reliable method for detecting fiducial points. Thereby, it provides a valuable new resource for researchers and healthcare professionals working in the analysis of photoplethysmography signals.

Photoplethysmography is a non-invasive technique used for measuring several vital signs and for the identification of individuals with an increased disease risk. Its principle of work is based on detecting changes in blood volume in the microvasculature of the skin through the absorption of light. The extraction of relevant features from the photoplethysmography signal for estimating certain physiological parameters is a challenging task, where various feature extraction methods have been proposed in the literature. In this work, we present PPGFeat, a novel MATLAB toolbox supporting the analysis of raw photoplethysmography waveform data. PPGFeat allows for the application of various preprocessing techniques, such as filtering, smoothing, and removal of baseline drift; the calculation of photoplethysmography derivatives; and the implementation of algorithms for detecting and highlighting photoplethysmography fiducial points. PPGFeat includes a graphical user interface allowing users to perform various operations on photoplethysmography signals and to identify, and if required also adjust, the fiducial points. Evaluating the PPGFeat’s performance in identifying the fiducial points present in the publicly available PPG-BP dataset, resulted in an overall accuracy of 99% and 3038/3066 fiducial points were correctly identified. PPGFeat significantly reduces the risk of errors in identifying inaccurate fiducial points. Thereby, it is providing a valuable new resource for researchers for the analysis of photoplethysmography signals.

In recent years, the field of prosthetics has made significant progress towards creating prosthetic devices that are more functional, comfortable, and user-friendly. However, achieving intuitive control over prosthetic hand movements remains a significant challenge, especially for individuals with limb loss who rely on prosthetics for independent daily activities. To address this challenge, researchers have explored the potential of non-invasive techniques as electromyography (EMG) for prosthetic control. This paper aims to investigate the potential of using EMG and the electrical impedance myography (EIMG) techniques jointly for the measurement of hand movements. The study involved recording and comparing EMG and EIMG signals from a cohort of healthy individuals. These signals were captured during four distinct hand gestures: opening and closing the hand, as well as extending and flexing it, under varying time conditions, allowing for categorization into low and high-intensity movements. Data collection employed the Open BCI and ZRPI devices. The analysis of these signal waveforms revealed compelling results. Brachioradialis activity in EMG 2 exhibited an increase during open hand (0.015mV) and extension hand (0.009mV in low and 0.013mV in high intensity) gestures, accompanied by increased EIMG activity (56mV and 52mV respectively). Additionally, close hand (0.0018mV in low and 0.05mV in high intensity) and flexion hand (0.0075 in low intensity and 0.002 in high intensity) gestures exhibited heightened flexor carpi ulnaris activity with raised EIMG activity (57mV and 45mV respectively). These results proved to be consistent, acceptable, and aligned with existing literature. The findings of this paper indicate that both EMG and EIMG techniques could be used together to control custom-made hand prosthetics, demonstrating a significant development that could lead to more intuitive and easier-to-control prosthetics. Also, the results obtained could be valuable to researchers and engineers working in the prosthetics field, as it provides insights into the potential of non-invasive techniques for prosthetic control.

Administering the medications and fluids intravenously is a frequent practice in modern medical procedures, which plays a vital role in the treatment of certain acute conditions which require immediate action by drugs or fluids. This paper covers the design of a low-cost, wireless drip monitoring system for use in the hospital environment. The device is equipped with the Bluetooth low energy based battery-operated microcontroller, an infrared based drops counting system and a digital servo motor to control the drip flow rate, and it is attached to an existing intravenous stand. A LabView graphical user interface has also been developed to provide sets of input to the system to calculate the desired drip rate and the amount of pressure that digital servo motor must apply to achieve it. The system shows an average accuracy of 96% when compared with the measured and calculated values. This allows accurate computation of the level of the drip.

Cardiovascular diseases (CVDs) are a leading cause of death worldwide, with hypertension emerging as a significant risk factor. Early detection and treatment of hypertension can significantly reduce the risk of developing CVDs and related complications. This work proposes a novel approach employing features extracted from the acceleration photoplethysmography (APG) waveform, alongside clinical parameters, to estimate different stages of hypertension. The current study used a publicly available dataset and a novel feature extraction algorithm to extract APG waveform features. Three distinct supervised machine learning algorithms were employed in the classification task, namely: Decision Tree (DT), Linear Discriminant Analysis (LDA), and Linear Support Vector Machine (LSVM). Results indicate that the DT model achieved exceptional training accuracy of 100% during cross-validation and maintained a high accuracy of 96.87% on the test dataset. The LDA model demonstrated competitive performance, yielding 85.02% accuracy during cross-validation and 84.37% on the test dataset. Meanwhile, the LSVM model exhibited robust accuracy, achieving 88.77% during cross-validation and 93.75% on the test dataset. These findings underscore the potential of APG analysis as a valuable tool for clinicians in estimating hypertension stages, supporting the need for early detection and intervention. This investigation not only advances hypertension risk assessment but also advocates for enhanced cardiovascular healthcare outcomes.

Today, well-established hazard analysis techniques are available and widely used to identify hazards for single systems in various industries. However, hazard analysis techniques for a System of Systems (SoS) are not properly investigated. SoS is a complex system where multiple systems work together to achieve a common goal. However, the interaction between systems may lead to unforeseen interactions and interdependencies between systems. This increases the difficulty of identifying and assessing system failures and potential safety hazards. In this paper, we explore whether Hazard Ontology (HO) can be applied to an SoS and whether it can identify emergent hazards, their causes, sources, and consequences. To conduct our exploration, we apply the HO to a quarry automation site (an SoS) from the construction equipment domain. The results indicate that the HO is a promising technique that facilitates the identification of emergent hazards and their components. 

While ontology comparison and alignment have been extensively researched in the last decade, there are still some challenges to these disciplines, such as incomplete ontologies, those that cover only a portion of a domain, and differences in domain modeling due to varying viewpoints. Although the literature has compared ontological concepts from the same domain, comparisons of concepts from different domains (e.g., security and safety) remain unexplored. To compare the concepts of security and safety domains, a security ontology must first be created to bridge the gap between these domains. Therefore, this paper presents a Combined Security Ontology (CSO) based on the Unified Foundational Ontology (UFO) that could be compared to or aligned with other ontologies. This CSO includes the core ontological concepts and their respective relationships that had been extracted through a previous systematic literature review. The CSO concepts and their relationships were mapped to the UFO to get a common terminology that facilitates to bridge the gap between the security and safety domains. Since the proposed CSO is based on the UFO, it could be compared to or aligned with other ontologies from different domains.

Safety and security ontologies quickly become essential support for integrating heterogeneous knowledge from various sources. Today, there is little standardization of ontologies and almost no discussion of how to compare concepts and their relationships, establish a general approach to create relationships or model them in general. However, concepts with similar names are not semantically similar or compatible in some cases. In this case, the problem of correspondence arises among the concepts and relationships found in the ontologies. To solve this problem, a comparison between the Hazard Ontology (HO) and the Combined Security Ontology (CSO) is proposed, in which the value of equivalence between their concepts and their relationships was extracted and analyzed. Although the HO covers the concepts related to the safety domain and the CSO includes securityrelated concepts, both are based on the Unified Foundational Ontology (UFO). For this study, HO and CSO were compared, and the results were summarized in the form of comparison tables. Our main contribution involves the comparisons among the concepts in HO and CSO to identify equivalences and differences between the two. Due to the increasing number of ontologies, their mapping, merging, and alignment are primary challenges in bridging the gaps that exist between the safety and security domains. 

Security ontologies have been developed to facilitate the organization and management of security knowledge. A comparison and evaluation of how these ontologies relate to one another is challenging due to their structure, size, complexity, and level of expressiveness. Differences between ontologies can be found on both the ontological and linguistic levels, resulting in errors and inconsistencies (i.e., different concept hierarchies, types of concepts, definitions) when comparing and aligning them. Moreover, many concepts related to security ontologies have not been thoroughly explored and do not fully meet security standards. By using standards, we can ensure that concepts and definitions are unified and coherent. In this study, we address these deficiencies by reviewing existing security ontologies to identify core concepts and relationships. The primary objective of the systematic literature review is to identify core concepts and relationships that are used to describe security issues. We further analyse and map these core concepts and relationships to five security standards (i.e., NIST SP 800-160, NIST SP 800-30 rev.1, NIST SP 800-27 rev.A, ISO/IEC 27001 and NISTIR 8053). As a contribution, this paper provides a set of core concepts and relationships that comply with the standards mentioned above and allow for a new security ontology to be developed.

With the development towards Industry 5.0, manufacturing companies are developing towards Smart Production, i.e., using data as a resource to interconnect the elements in the production system to learn and adapt accordingly for a more resource-efficient and sustainable production. This requires selecting and integrating digital technologies for the entire data lifecycle, also referred to as the data value chain. However, manufacturing companies are facing many challenges related to building data value chains to achieve the desired benefits of Smart Production. Therefore, the purpose of this paper is to identify and analyze the critical factors of selecting and integrating digital technologies for efficiently benefiting data value chains for Smart Production. This paper employed a qualitative-based multiple case study design involving manufacturing companies within different industries and of different sizes. The paper also analyses two Smart Production cases in detail by mapping the data flow using a technology selection and integration framework to propose solutions to the existing challenges. By analyzing the two in-depth studies and additionally two reference cases, 13 themes of critical factors for selecting and integrating digital technologies were identified.

To stay competitive, manufacturing companies are developing towards Smart Production which requires the use of digital technologies. However, there is a lack of guidance supporting manufacturing companies in selecting and integrating a combination of suitable digital technologies, which is required for Smart Production. To address this gap, the purpose of this paper is twofold: (i) to identify the main challenges of selecting and integrating digital technologies for Smart Production, and (ii) to propose a holistic concept to support manufacturing companies in mitigating identified challenges in order to select and integrate a combination of digital technologies for Smart Production. This is accomplished by using a qualitative-based multiple case study design. This paper identifies current challenges related to selection and integration of digital technologies. To overcome these challenges and achieve Smart production, the concept of data value chain was proposed, i.e., a holistic approach to systematically map and improve data flows within the production system. © 2022, IFIP International Federation for Information Processing.

Wireless sensors and networks are used only occasionally in current control loops in the process industry. With rapid developments in embedded and highperformance computing, wireless communication, and cloud technology, drastic changes in the architecture and operation of industrial automation systems seem more likely than ever. These changes are driven by ever-growing demands on production quality and flexibility. However, as discussed in "Summary," there are several research obstacles to overcome. The radio communication environment in the process industry is often troublesome, as the environment is frequently cluttered with large metal objects, moving machines and vehicles, and processes emitting radio disturbances [1], [2]. The successful deployment of a wireless control system in such an environment requires careful design of communication links and network protocols as well as robust and reconfigurable control algorithms.

This thesis presents a distributed, collaborative simultaneous localization and mapping (C-SLAM) approach for embedded systems aiming to support search and rescue missions. A distributed C-SLAM system provides the robustness and flexibility to explore unpredictable environments reliably, even in the case of an unexpected loss of one or multiple of the robots. The approach is demonstrated on the Turtlebot4 and uses ROS2 as the development platform and assumes known initial positions of the robots for the map merging process. The robots in the C-SLAM system constantly share their progress to guarantee the workload can be distributed across all robots and a task can be completed even if individual robots fail. Our proposed system's efficiency is measured as the time needed to explore a given unknown environment. As a result, the C-SLAM system achieves an average of approximately 40 % improvement compared to a single robot system, even with the loss of robots during exploration. This emphasizes the robustness of the system but also the relevance and need for a better navigation system to take full advantage of the multi-robot nature of a C-SLAM system.

Many test prioritisation techniques have been proposed in order to improve test effectiveness of Continuous Integration (CI) pipelines. Particularly, diversity-based testing (DBT) has shown promising and competitive results to improve test effectiveness. We report on a case study considering the CI pipeline of Axis Communications in Sweden. We compared three different prioritisation approaches (i.e., diversity, failure history and time) in terms of their impact on coverage, failure detection rates and reduction on test execution time. Our results reveal that DBT is the best candidate to provide feature coverage, whereas failure rate prioritisation yields better failure coverage. Time-based prioritisation is not a reliable approach to provide cost-effective testing. Moreover, DBT would allow stakeholders to receive quick feedback on many combinations of integrated features to verify their code changes. Our participants report that developers are mainly interested in: (i) receiving quick feedback on a high combination of integrated features to verify their code changes, and (ii) associate their test suites to confidence scores representing the risk of missing failures given that fewer tests are executed.

Convolutional Neural Networks (CNNs) have become integral in safety-critical applications, thus raising concerns about their fault tolerance. Conventional hardwaredependent fault tolerance methods, such as Triple Modular Redundancy (TMR), are computationally expensive, imposing a remarkable overhead on CNNs. Whereas fault tolerance techniques can be applied either at the hardware level or at the model levels, the latter provides more flexibility without sacrificing generality. This paper introduces a model-level hardening approach for CNNs by integrating error correction directly into the neural networks. The approach is hardwareagnostic and does not require any changes to the underlying accelerator device. Analyzing the vulnerability of parameters enables the duplication of selective filters/neurons so that their output channels are effectively corrected with an efficient and robust correction layer. The proposed method demonstrates fault resilience nearly equivalent to TMR-based correction but with significantly reduced overhead. Nevertheless, there exists an inherent overhead to the baseline CNNs. To tackle this issue, a cost-effective parameter vulnerability based pruning technique is proposed that outperforms the conventional pruning method, yielding smaller networks with a negligible accuracy loss. Remarkably, the hardened pruned CNNs perform up to 24% faster than the hardened un-pruned ones.

The reliability of Artificial Neural Networks (ANNs) has emerged as a prominent research topic due to their increasing utilization in safety-critical applications. Long Short-Term Memory (LSTM) ANNs have demonstrated significant advantages in healthcare applications, primarily attributed to their robust processing of time-series data and memory-facilitated capabilities. This paper, for the first time, presents a comprehensive and fine-grain analysis of the resilience of LSTM-based ANNs in the context of gait analysis using fault injection into weights. Additionally, we improve their resilience by replacing faulty weights with zero, enabling ANNs to withstand environments that are up to 20 times harsher while experiencing up to 7 times fewer critical faults than an unprotected ANN.

Artificial Intelligence (AI) and, in particular, Machine Learning (ML), have emerged to be utilized in various applications due to their capability to learn how to solve complex problems. Over the past decade, rapid advances in ML have presented Deep Neural Networks (DNNs) consisting of a large number of neurons and layers. DNN Hardware Accelerators (DHAs) are leveraged to deploy DNNs in the target applications. Safety-critical applications, where hardware faults/errors would result in catastrophic consequences, also benefit from DHAs. Therefore, the reliability of DNNs is an essential subject of research. In recent years, several studies have been published accordingly to assess the reliability of DNNs. In this regard, various reliability assessment methods have been proposed on a variety of platforms and applications. Hence, there is a need to summarize the state-of-the-art to identify the gaps in the study of the reliability of DNNs. In this work, we conduct a Systematic Literature Review (SLR) on the reliability assessment methods of DNNs to collect relevant research works as much as possible, present a categorization of them, and address the open challenges. Through this SLR, three kinds of methods for reliability assessment of DNNs are identified, including Fault Injection (FI), Analytical, and Hybrid methods. Since the majority of works assess the DNN reliability by FI, we characterize different approaches and platforms of the FI method comprehensively. Moreover, Analytical and Hybrid methods are propounded. Thus, different reliability assessment methods for DNNs have been elaborated on their conducted DNN platforms and reliability evaluation metrics. Finally, we highlight the advantages and disadvantages of the identified methods and address the open challenges in the research area. We have concluded that Analytical and Hybrid methods are light-weight yet sufficiently accurate and have the potential to be extended in future research and to be utilized in establishing novel DNN reliability assessment frameworks.

The superior performance of Deep Neural Networks (DNNs) has led to their application in various aspects of human life. Safety-critical applications are no exception and impose rigorous reliability requirements on DNNs. Quantized Neural Networks (QNNs) have emerged to tackle the complexity of DNN accelerators, however, they are more prone to reliability issues.In this paper, a recent analytical resilience assessment method is adapted for QNNs to identify critical neurons based on a Neuron Vulnerability Factor (NVF). Thereafter, a novel method for splitting the critical neurons is proposed that enables the design of a Lightweight Correction Unit (LCU) in the accelerator without redesigning its computational part.The method is validated by experiments on different QNNs and datasets. The results demonstrate that the proposed method for correcting the faults has a twice smaller overhead than a selective Triple Modular Redundancy (TMR) while achieving a similar level of fault resiliency. 

Deep Neural Networks (DNNs) and their accelerators are being deployed ever more frequently in safety-critical applications leading to increasing reliability concerns. A traditional and accurate method for assessing DNNs' reliability has been resorting to fault injection, which, however, suffers from prohibitive time complexity. While analytical and hybrid fault injection-/analyticalbased methods have been proposed, they are either inaccurate or specific to particular accelerator architectures. In this work, we propose a novel accurate, fine-grain, metric-oriented, and accelerator-agnostic method called DeepVigor that provides vulnerability value ranges for DNN neurons' outputs. An outcome of DeepVigor is an analytical model representing vulnerable and non-vulnerable ranges for each neuron that can be exploited to develop different techniques for improving DNNs' reliability. Moreover, DeepVigor provides reliability assessment metrics based on vulnerability factors for bits, neurons, and layers using the vulnerability ranges. The proposed method is not only faster than fault injection but also provides extensive and accurate information about the reliability of DNNs, independent from the accelerator. The experimental evaluations in the paper indicate that the proposed vulnerability ranges are 99.9% to 100% accurate even when evaluated on previously unseen test data. Also, it is shown that the obtained vulnerability factors represent the criticality of bits, neurons, and layers proficiently. DeepVigor is implemented in the PyTorch framework and validated on complex DNN benchmarks.

Deep Learning, and in particular, Deep Neural Network (DNN) is nowadays widely used in many scenarios, including safety-critical applications such as autonomous driving. In this context, besides energy efficiency and performance, reliability plays a crucial role since a system failure can jeopardize human life. As with any other device, the reliability of hardware architectures running DNNs has to be evaluated, usually through costly fault injection campaigns. This paper explores approximation and fault resiliency of DNN accelerators. We propose to use approximate (AxC) arithmetic circuits to agilely emulate errors in hardware without performing fault injection on the DNN. To allow fast evaluation of AxC DNN, we developed an efficient GPU-based simulation framework. Further, we propose a fine-grain analysis of fault resiliency by examining fault propagation and masking in networks.

Trigger-Action Platforms (TAPs) empower applications (apps) for connecting otherwise unconnected devices and services. The current TAPs like IFTTT require trigger services to push excessive amounts of sensitive data to the TAP regardless of whether the data will be used in the app, at odds with the principle of data minimization. Furthermore, the rich features of modern TAPs, including IFTTT queries to support multiple trigger services and nondeterminism of apps, have been out of the reach of previous data minimization approaches like minTAP. This paper proposes LazyTAP, a new paradigm for fine-grained on-demand data minimization. LazyTAP breaks away from the traditional push-all approach of coarse-grained data over-approximation. Instead, LazyTAP pulls input data on-demand, once it is accessed by the app execution. Thanks to the fine granularity, LazyTAP enables tight minimization that naturally generalizes to support multiple trigger services via queries and is robust with respect to nondeterministic behavior of the apps. We achieve seamlessness for third-party app developers by leveraging laziness to defer computation and proxy objects to load necessary remote data behind the scenes as it becomes needed. We formally establish the correctness of LazyTAP and its minimization properties with respect to both IFTTT and minTAP. We implement and evaluate LazyTAP on app benchmarks showing that on average LazyTAP improves minimization by 95% over IFTTT and by 38% over minTAP, while incurring a tolerable performance overhead. 

Trigger-Action Platforms (TAPs) enable applications to integrate various devices and services otherwise unconnected. Recent features of TAPs introduce additional sources of data such as queries in IFTTT. The current TAPs, like IFTTT, demand that trigger and query services transmit excessive amounts of user data to the TAP. To limit the data to what is actually necessary for the execution to comply with the principle of data minimization, input services should send no more than the necessary data. LazyTAP proposes a new paradigm of data minimization by construction in TAPs, introducing a novel perspective for data collection from input services. While the existing push-all approach of TAPs entails coarse-grained data over-approximation, LazyTAP pulls input data on-demand at the level of attributes, once accessed by the app execution. Thanks to the fine granularity provided by LazyTAP, multiple trigger and query services can be naturally minimized while the behavior of app executions is preserved. In addition, a great benefit of LazyTAP is being seamless for third-party app developers. By leveraging laziness, LazyTAP defers computation and proxies objects to load necessary remote data behind the scenes. Our evaluation study on app benchmarks shows that on average LazyTAP improves minimization by 95% over IFTTT and by 38% over minTAP, with a tolerable performance overhead. This poster goes into further details about LazyTAP and elaborates on its prototype implementation. 

The aim of this work is to apply and evaluate different chemometric approaches employing several machine learning techniques in order to characterize the moisture content in biomass from data obtained by Near Infrared (NIR) spectroscopy. The approaches include three main parts: a) data pre-processing, b) wavelength selection and c) development of a regression model enabling moisture content measurement. Standard Normal Variate (SNV), Multiplicative Scatter Correction and Savitzky-Golay first (SGi) and second (SG2) derivatives and its combinations were applied for data pre-processing. Genetic algorithm (GA) and iterative PLS (iPLS) were used for wavelength selection. Artificial Neural Network (ANN), Gaussian Process Regression (GPR), Support Vector Regression (SVR) and traditional Partial Least Squares (PLS) regression, were employed as machine learning regression methods. Results shows that SNV combined with SG1 first derivative performs the best in data pre-processing. The GA is the most effective methods for variable selection and GPR achieved a high accuracy in regression modeling while having low demands on computation time. Overall, the machine learning techniques demonstrate a great potential to be used in future NIR spectroscopy applications.

AI methods referred to as interpretable are often discredited as inaccurate by supporters of the existence of a trade-off between interpretability and accuracy. In many problem contexts however this trade-off does not hold. This paper discusses a regression problem context to predict flight take-off delays where the most accurate data regression model was trained via the XGBoost implementation of gradient boosted decision trees. While building an XGB-CBR Twin and converting the XGBoost feature importance into global weights in the CBR model, the resultant CBR model alone provides the most accurate local prediction, maintains the global importance to provide a global explanation of the model, and offers the most interpretable representation for local explanations. This resultant CBR model becomes a benchmark of accuracy and interpretability for this problem context, and hence it is used to evaluate the two additive feature attribute methods SHAP and LIME to explain the XGBoost regression model. The results with respect to local accuracy and feature attribution lead to potentially valuable future work. © 2022 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). CEUR Workshop Proceedings (CEUR-WS.org)

Proactive maintenance management of world-class standard is close to impossible without the support of a computerized management system. In order to reduce failures, and failure recurrence, the key information to log are failure causes. However, Computerized Maintenance Management System (CMMS) seems to be scarcely used for analysis for improvement initiatives. One part of this is due to the fact that many CMMS utilizes free-text fields which may be difficult to analyze statistically. The aim of this study is to apply Natural Language Processing (NPL), Ontology and Machine Learning (ML) as a means to analyze free-textual information from a CMMS. Through the initial steps of the study, it was concluded though that none of these methods were able to find any suitable hidden patterns with high-performance accuracy that could be related to recurring failures and their root causes. The main reason behind that was that the free-textual information was too unstructured, in terms of for instance: spelling- and grammar mistakes and use of slang. That is the quality of the data are not suitable for the analysis. However, several improvement potentials in reporting and to develop the CMMS further could be provided to the company so that they in the future more easily will be able to analyze its maintenance data.

This report presents a research study plan on human subjects – the influence of stress and alcohol in simulated traffic situations under an H2020 project named SIMUSAFE. This research study focuses on road-users’, i.e., car drivers, motorcyclists, bicyclists and pedestrians, behaviour in relation to retrospective studies, where interaction between the users are considered. Here, the study includes sample size, inclusion/exclusion criteria, detailed study plan, protocols, potential test scenarios and all related ethical issues. The study plan has been included in a national ethics application and received approval for implementation.

Safety is a fundamental prerequisite that must be addressed before any interaction of robots with humans. Safety has been generally understood and studied as the physical safety of robots in human–robot interaction, whereas how humans perceive these robots has received less attention. Physical safety is a necessary condition for safe human–robot interaction. However, it is not a sufficient condition. A robot that is safe by hardware and software design can still be perceived as unsafe. This article focuses on perceived safety in human–robot interaction. We identified six factors that are closely related to perceived safety based on the literature and the insights obtained from our user studies. The identified factors are the context of robot use, comfort, experience and familiarity with robots, trust, the sense of control over the interaction, and transparent and predictable robot actions. We then made a literature review to identify the robot-related factors that influence perceived safety. Based the literature, we propose a taxonomy which includes human-related and robot-related factors. These factors can help researchers to quantify perceived safety of humans during their interactions with robots. The quantification of perceived safety can yield computational models that would allow mitigating psychological harm.

Human hands and fingers have been widely studied in different fields, such as animation, biomechanics, ergonomics, rehabilitation, medicine, and robotics. However, since the hands are a highly complex part of the human body capable of developing precise tasks, replicating human hand mechanisms remains challenging and, thus, continues to be an active area. This study focuses on a bioinspired mechanically equivalent finger model. A parametric model was proposed based on the typical architecture of a human finger, allowing adaptation to different anthropometries. A forward kinematic model assesses each phalanx's range of motion (ROM) during flexion-extension and abduction-adduction. A CAD modeling technique based on fused filament fabrication (FFF) is used for easy fabrication, requiring no assembly. The resulting model achieves the desired ROM, offering a simple solution for hand modeling. This bioinspired model aids in training hand exoskeleton robots, accurately mimicking human finger mechanics for various applications, including rehabilitation and prosthetics. This model helps understand complex hand mechanisms and holds potential for robotics and related fields.

Utilizing edge and cloud computing to empower the profitability of manufacturing is drastically increasing in modern industries. As a result of that, several challenges have raised over the years that essentially require urgent attention. Among these, coping with different faults in edge and cloud computing and recovering from permanent and temporary faults became prominent issues to be solved. In this paper, we focus on the challenges of applying fault tolerance techniques on edge and cloud computing in the context of manufacturing and we investigate the current state of the proposed approaches by categorizing them into several groups. Moreover, we identify critical gaps in the research domain as open research directions. 

Cloud computing is a game-changer model that opens new directions for modern manufacturing. It enables services and solutions that help improve the productivity and efficiency of smart production plants. The main objective of the paper is to provide a summary of the various cloud-based manufacturing services currently being offered to manufacturers or that could be offered in the future. Additionally, the paper aims to discuss the various enabling technologies used to support the integration of cloud manufacturing in the manufacturing industry. Furthermore, the paper categorizes the different services based on their functionalities and maps them to four levels of production such as plant level, production line level, machine level, and process level. The categorization of services and mapping them to appropriate levels in production can enhance efficiency and productivity in the manufacturing industry. The study advances the discussion on cloud-based manufacturing from the types of services and enabling technologies perspective.

In the era of the IoT revolution, applications are becoming ever more sophisticated and accompanied by diverse functional and non-functional requirements, including those related to computing resources and performance levels. Such requirements make the development and implementation of these applications complex and challenging. Computing models, such as cloud computing, can provide applications with on-demand computation and storage resources to meet their needs. Although cloud computing is a great enabler for IoT and endpoint devices, its limitations make it unsuitable to fulfill all design goals of novel applications and use cases. Instead of only relying on cloud computing, leveraging and integrating resources at different layers (like IoT, edge, and cloud) is necessary to form and utilize a computing continuum. The layers’ integration in the computing continuum offers a wide range of innovative services, but it introduces new challenges (e.g., monitoring performance and ensuring security) that need to be investigated. A better grasp and more profound understanding of the computing continuum can guide researchers and developers in tackling and overcoming such challenges. Thus, this paper provides a comprehensive and unified view of the computing continuum. The paper discusses computing models in general with a focus on cloud computing, the computing models that emerged beyond the cloud, and the communication technologies that enable computing in the continuum. In addition, two novel reference architectures are presented in this work: one for edge–cloud computing models and the other for edge–cloud communication technologies. We demonstrate real use cases from different application domains (like industry and science) to validate the proposed reference architectures, and we show how these use cases map onto the reference architectures. Finally, the paper highlights key points that express the authors’ vision about efficiently enabling and utilizing the computing continuum in the future.