Life-critical Applications Research Articles

Navigating ethical challenges of explainable ai in autonomous systems

The rapid integration of autonomous systems, such as vehicles, drones, and robots, into various sectors brings forth significant ethical challenges concerning their decision-making processes. This paper examines the role of Explainable AI (XAI) in addressing these challenges, particularly regarding accountability in the event of accidents and the necessity of human oversight in automated environments. We discuss the critical ethical implications of transparency, emphasizing how XAI can bridge the gap between complex algorithmic decision-making and public understanding, thereby fostering trust in these technologies. The paper also outlines the current regulatory frameworks for AI safety, analysing their effectiveness in promoting responsible innovation. Furthermore, we investigate the consequences of opaque algorithms, especially in life-critical applications where the stakes are exceptionally high. Through an analysis of case studies, we showcase how organizations have successfully implemented XAI to enhance safety measures and uphold ethical responsibility in their autonomous systems. Ultimately, this study advocates for the integration of XAI as a vital component in developing responsible autonomous technologies, ensuring accountability, and safeguarding public trust in an era increasingly defined by automation.

Toward Fast and Energy-Efficient Access to Cloudlets in Hostile Environments

Cloudlets, which refer to the edge computing services deployed at the proximity of end devices, are key providers of connectivity, storage, and computation resources to many applications. While access to cloudlets is pervasive in typical settings, it can be difficult in challenging, even hostile environments, such as military or post-disaster scenarios, featuring multi-hop communication and energy-constrained end devices. In these cases, cloudlets may have become the only equipment powerful enough to execute life-critical applications, such as battle-field situation awareness, tactic cooperation, and search-and-rescue missions. Quality of these services is greatly influenced by the minimum time that a packet can be delivered, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">i.e</i> ., the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">cloudlet access delay</i> (CAD), whose characteristics remain unknown. To address the open question of fast and efficient cloudlet access, we establish a <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">packet mobility</i> model that allows CAD and energy consumption to be analyzed as a function of the initial device-cloudlet distance. We find that the expected CAD scales either linearly or quadratically under distinct types of packet mobility, and the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">successful access rate</i> (SAR) can be bounded by functions of the delay constraint. Based on these findings, we develop a packet shedding algorithm that saves 24% transmission power, and reduces the average CAD by 2%, while maintaining a similar SAR in simulated cloudlet access environments.

Enhancing Semantic Web Technologies Using Lexical Auditing Techniques for Quality Assurance of Biomedical Ontologies

Semantic web technologies (SWT) represent data in a format that is easier for machines to understand. Validating the knowledge in data graphs created using SWT is critical to ensure that the axioms accurately represent the so-called “real” world. However, data graph validation is a significant challenge in the semantic web domain. The Shapes Constraint Language (SHACL) is the latest W3C standard developed with the goal of validating data-graphs. SHACL (pronounced as shackle) is a relatively new standard and hitherto has predominantly been employed to validate generic data graphs like WikiData and DBPedia. In generic data graphs, the name of a class does not affect the shape of a class, but this is not the case with biomedical ontology data graphs. The shapes of classes in biomedical ontology data graphs are highly influenced by the names of the classes, and the SHACL shape creation methods developed for generic data graphs fail to consider this characteristic difference. Thus, the existing SHACL shape creation methods do not perform well for domain-specific biomedical ontology data graphs. Maintaining the quality of biomedical ontology data graphs is crucial to ensure accurate analysis in safety-critical applications like Electronic Health Record (EHR) systems referencing such data graphs. Thus, in this work, we present a novel method to create enhanced SHACL shapes that consider the aforementioned characteristic difference to better validate biomedical ontology data graphs. We leverage the knowledge available from lexical auditing techniques for biomedical ontologies and incorporate this knowledge to create smart SHACL shapes. We also create SHACL shapes (baseline SHACL graph) without incorporating the lexical knowledge of the class names, as is performed by existing methods, and compare the performance of our enhanced SHACL shapes with the baseline SHACL shapes. The results demonstrate that the enhanced SHACL shapes augmented with lexical knowledge of the class names identified 176 violations which the baseline SHACL shapes, void of this lexical knowledge, failed to detect. Thus, the enhanced SHACL shapes presented in this work significantly improve the validation performance of biomedical ontology data graphs, thereby reducing the errors present in such data graphs and ensuring safe use in the life-critical applications referencing them.

Novel MAC protocol for handling correlation and dynamics of medical traffic in WBANs

In Wireless Body Area Networks (WBANs), it is crucial to ensure reliable and timely delivery of gathered data, despite their intrinsic challenges related to limited resources at the sensor level. In fact, in the medical context, most real-time and life-critical applications are both delay-sensitive and loss-sensitive; the loss or corruption packets carrying critical data due to unreliable wireless networks could have severe consequences. In this regard, the concept of context awareness can be exploited; the more an application acquires additional information about users, sensors, and their context, the better it can adapt its functional behavior to improve its performance of delay, throughput and packet loss.Thus, acquiring some knowledge about the patient’s medical state over time may enhance the interpretation and intervention of the medical staff. In this context, we propose a dynamic context-aware MAC protocol (DMTM-MAC) that takes into account the dynamic and correlation aspects of medical traffic in order to adjust the channel access mechanisms and the general network behavior. We assign to each sensor a context that reflects the type of traffic generated and its QoS constraints. We aim to track specific biosensors, which are related to vital indicators relevant to a specific disease; the associated measurements may reflect the variation of the patient’s state over time. To such nodes, we assign a new node context, called dynamic context, which can help to predict the eventual health decline over time. The body node coordinator will have the ability to analyze the received data in order to detect abnormalities and disease evolution. Observed changes in collected measurements can be an early predictor of an advanced or complicated situation and can allow this BAN coordinator to handle dynamic context changes in these types of sensors by increasing their priority and transmission frequency to cope with the current situation. Performance evaluation showed that DMTM-MAC outperforms both MTM-MAC and IEEE 802.15.6 in terms of delay and packet delivery ratio. The proposed approach enhances the delay up to 80 % over IEEE 802.15.6.

Understanding resistance increase in composite inks under monotonic and cyclic stretching

Cyclic degradation in flexible electronic inks remains a key challenge while their deployment in life critical applications is ongoing. The origin of electrical degradation of a screen-printed stretchable conductive ink with silver flakes embedded in a polyurethane binder is investigated under uniaxial monotonic and cyclic stretching, using in-situ confocal microscopy and scanning electron microscopy experiments, for varying ink thickness (1, 2, and 3 layers, each layer around 8–10 μm) and trace width (0.5, 1, and 2 mm). Cracks form under monotonic stretching, and the evolution of crack pattern (density, length and width) with applied strain is affected by ink thickness such that the 3-layer ink exhibits larger normalized resistance but slightly lower resistance than the 1-layer ink up to strains of 125%. For cyclic stretching, the crack density and length do not evolve with cycling. However, the cracks widen and deepen, leading to an increase in resistance with cycling. There exists a strong correlation between fatigue life, i.e. the number of cycles until a normalized resistance of 100 is reached, and the strain amplitude. The normalized resistance increase rate with respect to cycling is also found to scale with strain amplitude. The rate of change in resistance with cycling decreases with ink thickness and trace width. For practical applications, thicker (25 μm) and wider (⩾2 mm) inks should be used to lower resistance increases with repeated deformation.

Collaborative Defense-GAN for protecting adversarial attacks on classification system

With rapid progress and significant successes in a wide domain of applications, deep learning has been extensively employed for solving complex problems. However, performance of deep learning has been vulnerable to well-designed samples, called adversarial samples. These samples are carefully designed to deceive the deep learning models without human perception. Therefore, vulnerability to adversarial attacks becomes one of the major concerns in life-critical applications of deep learning. In this paper, a novel approach to counter adversarial samples is proposed to strengthen the robustness of a deep learning model. The strategy is to filter the perturbation noise in adversarial samples prior to prediction. The proposed defense framework is based on DiscoGANs to discover the relation between attacker and defender characteristics. Attacker models are created to generate the adversarial samples from the training data, while the defender model is trained to reconstruct original samples from the adversarial samples. These two frameworks are trained to compete with each other in an alternating manner. The experimental results on different attack models are compared with popular defense mechanisms on three benchmark datasets. Our proposed method shows promising results and can improve the robustness on both white-box and black-box attacks including the computation time.

A regularization method to improve adversarial robustness of neural networks for ECG signal classification

With the advancement of machine leaning technologies, Deep Neural Networks (DNNs) have been utilized for automated interpretation of Electrocardiogram (ECG) signals to identify potential abnormalities in a patient's heart within a second. Studies have shown that the accuracy of DNNs for ECG signal classification could reach human-expert cardiologist level if a sufficiently large training dataset is available. However, it is known that, in the field of computer vision, DNNs are not robust to adversarial noises that may cause DNNs to make wrong class-label predictions. In this work, we confirm that DNNs are not robust to adversarial noises in ECG signal classification applications, and we propose a novel regularization method to improve DNN robustness by minimizing the noise-to-signal ratio. Our method is evaluated on two public datasets: the MIT-BIH dataset and the CPSC2018 dataset, and the evaluation results show that our method can significantly enhance DNN robustness against adversarial noises generated by Projected Gradient Descent (PGD) and Smooth Adversarial Perturbation (SAP) adversarial attacks, with a minimal reduction of accuracy on clean data. Our method may serve as the baseline for designing new methods to defend against adversarial attacks for life-critical applications depending on ECG interpretation. The code of this work is publicly available at github.com/SarielMa/Robust_DNN_for_ECG.

Evaluation of Smartphone’s Embedded Sensors Through Applications: A Case Study of Gyroscope and Accelerometer

Smartphones are used for many daily activities like tele-communication, gaming, web browsing, fitness and health monitoring and traditional office working. Smartphones are equipped with built-in sensors to be able to perform these activities. It is well known that the sensors affect the resolution of the smartphone applications which is very vital in life critical applications (LCA). In this paper, two main sensors, the gyroscope and accelerometer have been studied. All commercial smartphones contain these two sensors and support functions related to them. These two sensors have direct link with the physical measurements which feed the fitness and health applications. A fitness application has been selected and ran under Android and iOS operating systems in two different popular smartphones: Samsung Note5 and iPhone7s smartphones. Statistical methodology has been applied to analysis the data and evaluate the performance of the sensors. The results show that commercial smartphones are not reliable devices for motion-related measurements and they can only be used for general purpose monitoring but not in life critical applications.

Drift Velocity Estimation of Ionospheric Bubbles Using GNSS Observations

AbstractEquatorial plasma bubble (EPB) irregularities are large‐scale plasma depleted structures that exist in the equatorial and low‐latitude ionosphere. Thus, particularly in these regions, much attention must be given to the effects of the EPBs in the life critical applications based on global navigation satellite systems (GNSS). The study on the dynamics of plasma bubbles, particularly in the determination of their drift velocities, and their impact on ground‐based augmentation system is of fundamental importance for civilian aviation. In this regard, we proposed a methodology to estimate the zonal drift velocities of the plasma bubble irregularities using slant total electron content (TEC) measurements derived from two spaced ground‐based GNSS receivers. The experiments were performed with the purpose to evaluate if the drift velocities obtained using the methodology proposed in this study agree with the velocities deduced from all‐sky imaging systems. The results revealed that the TEC‐estimated mean eastward drift velocities were comparable with the values deduced from the airglow techniques, and with the results obtained from previous studies. Therefore, the methodology proposed in this work to infer the plasma bubble drift velocities seem to have the potential in future investigations to provide a new data source in this field.

Globally Optimal Consensus Maximization for Relative Pose Estimation With Known Gravity Direction

Relative pose estimation is a core task in robotic vision, and it is the basis of many high-level applications (e.g., visual odometry). In this letter, we focus on a quite common case in which the gravity direction is known in advance with the help of IMUs. Commonly, incorrect feature matches (a.k.a. outliers) are unavoidable, and they will impair the accuracy significantly. RANSAC is the de facto standard to suppress the outliers and obtain a robust solution. However, RANSAC is a non-deterministic algorithm, which means it produces a reasonable result only with a certain probability, and it cannot guarantee the global optimality to meet the safety demand in many life-critical applications. Therefore, we propose a globally optimal algorithm for relative pose estimation with known gravity direction. Specifically, the proposed method employs the branch-and-bound algorithm to solve a consensus maximization problem, and thus it is able to obtain the global solution with a provable guarantee. To verify the feasibility of our proposed method, both synthetic and real-data experiments are conducted. The experimental results support the global optimality of the proposed method and show that the method performs more robustly than existing methods.

Defending Support Vector Machines Against Data Poisoning Attacks

Support Vector Machines (SVMs) are vulnerable to targeted training data manipulations such as poisoning attacks and label flips. By carefully manipulating a subset of training samples, the attacker forces the learner to compute an incorrect decision boundary, thereby causing misclassifications. Considering the increased importance of SVMs in engineering and life-critical applications, we develop a novel defense algorithm that improves resistance against such attacks. Local Intrinsic Dimensionality (LID) is a promising metric that characterizes the outlierness of data samples. In this work, we introduce a new approximation of LID called K-LID that uses kernel distance in the LID calculation, which allows LID to be calculated in high dimensional transformed spaces. We introduce a weighted SVM against such attacks using K-LID as a distinguishing characteristic that de-emphasizes the effect of suspicious data samples on the SVM decision boundary. Each sample is weighted on how likely its K-LID value is from the benign K-LID distribution rather than the attacked K-LID distribution. Experiments with benchmark data sets show that the proposed defense reduces classification error rates substantially (10% on average).

Fuzzy allocation model for health care data management on IoT assisted wearable sensor platform

In recent times, networking, information, and communication technologies are widely employed and getting spread in the area of medical applications. In a short period, real-time applications for healthcare monitoring produce a huge amount of data. Data is to be bursting in the case of critical applications; hence there is a greater need to enable reliable communication methods that ensure energy efficiency. A large amount of data results in high data transmission, network congestion, and latency, which in turn cause a hop increment between the IoT and cloud servers, and thus data might be unprocessed and insufficient for end-users. Since Fog computing is a distributed intermediate layer between the edge network and the cloud environment, latency can be reduced to a remarkable level, and the reliable communication can be achieved with the help of fog computing on IoT assisted wearable sensor platform. This research proposes a dynamic model for analysis and the Heuristic Hybrid Time Slot Fuzzy-Allocation Algorithm (HHTSF-AA), which improves health monitoring by indulging IoT assisted wearable sensor platform. Fog computing assisted wearable sensor platform is the most suitable and reliable platform for robust life-critical applications that are likely not to be delayed in communication. Besides, routing data packets are equipped with a low-cost energy minimum selection algorithm incorporated to improve the overall network performance. Dynamic slot assignment reduces time in a network and allows high levels of network capacity channel utilization.

Power-Management Strategies for Medical Information Transmission in Wireless Body Sensor Networks

To minimize and manage the power drain, and extend battery lifetime of wireless body sensor networks (WBSN) is one of the major challenges. There are three key purposes of this survey article, first, to examine the downsides of the classical power-management methods in WBSNs; second, considering the life-critical applications and emergency contexts that are encompassed by WBSN; and, third, studying the impact of power-management techniques on resource-confined networks for economical healthcare. A specific power-management solution is also discussed.

Towards 6G in-X Subnetworks With Sub-Millisecond Communication Cycles and Extreme Reliability

The continuous proliferation of applications requiring wireless connectivity will eventually result in latency and reliability requirements beyond what is achievable with current technologies. Such applications can for example include industrial control at the sensor-actuator level, intra-vehicle communication, fast closed loop control in intra-body networks and intra-avionics communication. In this article, we present the design of short range Wireless Isochronous Real Time (WIRT) in-X subnetworks aimed at life-critical applications with communication cycles shorter than 0.1 ms and outage probability below 10 -6 . Such targets are clearly beyond what is supported by the 5th Generation (5G) radio technology, and position WIRT as a possible 6th Generation (6G) system. WIRT subnetworks are envisioned to be deployed for instance in industrial production modules, robots, or inside vehicles. We identify technology components as well as spectrum bands for WIRT subnetworks and present major design aspects including frame structure and transmission techniques. The performance evaluation considering a dense scenario with up to 2 devices per m 2 reveal that a multi-GHz spectrum may be required for ensuring high spatial service availability. The possibility of running WIRT as an ultra-wideband underlay system in the centimeter-wave spectrum region is also discussed. Aspects related to design of techniques for the control plane as well as enhancements to the presented design is the focus of our ongoing research.

Fault-Tolerant Modular Stator Concentrated Winding Permanent Magnet Machine

This paper studies a modular permanent magnet synchronous machine (MPMSM) converted from a traditional single-core 24-slot 20-pole tooth-coil winding permanent magnet synchronous machine (TCW PMSM). The performance of the TCW PMSM is compared with the performance of the MPMSM (rearranged from this TCW PMSM) by the finite element method (FEM). It is found that if an electrical machine with a modular structure is designed, the efficiency of the MPMSM may be degraded compared with the conventional structure, especially, if a solid rotor yoke is used. However, the advantages of the proposed MPMSM are related to an option to scale up the machine by applying a certain number of modules, or removing faulty modules (while keeping the healthy modules working). This makes this type of a motor design applicable in certain life-critical applications, or in applications where scaling of the motor power might be needed. As a prototype, a low-power, low-speed generator added to a hook block of a hoist to supply power to measurement electronics is studied.

A Fault-Tolerant Scheduling Algorithm Based on Checkpointing and Redundancy for Distributed Real-Time Systems

Real-time systems are becoming ever more widely used in life-critical applications, and the need for fault-tolerant scheduling can only grow in the years ahead. This article presents a novel fault tolerance approach for tolerating transient faults in hard real-time systems. The proposed approach combines both checkpointing with rollback and active replication to tolerate several transient faults. Based on this approach, a new static fault-tolerant scheduling algorithm SFTS is presented. It is based on a list of scheduling heuristics which satisfy the application time constraints even in the presence of faults by exploring the spare capacity of available processors in the architecture. Simulation results show the performance and effectiveness of the proposed approach compared to other fault-tolerant approaches. The results reveal that in the presence of multiple transient faults, the average timing overhead of this approach is lower than checkpointing technique. Moreover, the proposed algorithm SFTS achieves better feasibility rate in the presence of multiple transient faults.

Fuzzy-Based Dynamic Time Slot Allocation forWireless Body Area Networks.

With the advancement in networking, information and communication technologies, wireless body area networks (WBANs) are becoming more popular in the field of medical and non-medical applications. Real-time patient monitoring applications generate periodic data in a short time period. In the case of life-critical applications, the data may be bursty. Hence the system needs a reliable energy efficient communication technique which has a limited delay. In such cases the fixed time slot assignment in medium access control standards results in low system performance. This paper deals with a dynamic time slot allocation scheme in a fog-assisted network for a real-time remote patient monitoring system. Fog computing is an extended version of the cloud computing paradigm, which is suitable for reliable, delay-sensitive life-critical applications. In addition, to enhance the performance of the network, an energy-efficient minimum cost parent selection algorithm has been proposed for routing data packets. The dynamic time slot allocation uses fuzzy logic with input variables as energy ratio, buffer ratio, and packet arrival rate. Dynamic slot allocation eliminates the time slot wastage, excess delay in the network and attributes a high level of reliability to the network with maximum channel utilization. The efficacy of the proposed scheme is proved in terms of packet delivery ratio, average end to end delay, and average energy consumption when compared with the conventional IEEE 802.15.4 standard and the tele-medicine protocol.

Influence of failure propagation on mission abort policy in heterogeneous warm standby systems

In many life-critical applications (e.g., spacecraft, aircrafts and chemical reactors), accomplishing a specified mission or aborting mission objectives in the case of certain condition being met to survive the system are both crucial. While conventional reliability models have typically focused only on evaluating the mission success probability, few of recent research addressed effects of mission aborts and studied system survivability. However, these works fail to consider propagated failures that can make significant contributions to the mission failure and may affect the optimal mission abort policy adopted for the system. In this paper, we suggest a numerical method for addressing effects of failure propagations in analyzing the mission success probability (MSP) and system survivability (SS) of 1-out-of-N warm standby systems subject to mission abort. The system may be heterogeneous, composed of components with different standby modes, and different performance and failure time distribution parameters. Both the activation sequence of those heterogeneous components and the adopted mission abort policy can affect the system performance metrics greatly. Therefore, we make further contributions by formulating and solving optimization problems that find the optimal mission abort policy for systems with any given component activation sequence or the optimal combination of mission abort policy and component activation sequence, maximizing the MSP while satisfying a desired SS level. Examples are provided to demonstrate influence of propagated failures on system performance and optimization solutions. Examples also show that the optimal abort policy determined can achieve a good balance between MSP and SS taking into account failure propagation.

Hybrid fuzzy multi-criteria decision making based multi cluster head dolphin swarm optimized IDS for VANET

Existing Vehicle-to-Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) based communication suffers from various security and performance issues, hence Cluster based Communication is preferred nowadays. However, Cluster based Communication adds extra overhead and burden on the Cluster Head (CH) in dense network scenarios which eventually introduces delay and hinders network performance. To reduce the overburdening of single CH, a multi cluster head scheme is proposed in which multiple nodes in a cluster can act as CH to share the load of single CH. For a selection of stable CH, Hybrid Fuzzy Multi-criteria Decision making approach (HF-MCDM) is proposed in which Fuzzy Analytic Hierarchy Process (AHP) and TOPSIS methods are clubbed together for optimal decision making. Further because of association of Vehicular Ad-hoc Network (VANET) with life-critical applications, there is a dire need for a security framework to detect various malevolent attacks. Machine Learning based Intrusion Detection System (IDS) like Support Vector Machine (SVM) is one of the approaches for curbing such attacks. These intrusion detection based mechanism can be combined with various existing optimization techniques to improve their performance, and Dolphin Swarm Algorithm is one such approach. Dolphins have many significant biological features like echolocation, exchange of information, coordination, and division of labor. These biological features combined with swarm intelligence can be utilized for optimizing the detection and accuracy of SVM based IDS. So in this paper, a Multi-Cluster Head anomaly based IDS optimized by Dolphin Swarm Algorithm has been proposed and its results are compared with various existing Security frameworks in terms of parameters like false positive, detection rate, detection time, etc. and it is observed that the proposed approach performs better.

A Novel Dual Separate Paths (DSP) Algorithm Providing Fault-Tolerant Communication for Wireless Sensor Networks.

Fault tolerance has long been a major concern for sensor communications in fault-tolerant cyber physical systems (CPSs). Network failure problems often occur in wireless sensor networks (WSNs) due to various factors such as the insufficient power of sensor nodes, the dislocation of sensor nodes, the unstable state of wireless links, and unpredictable environmental interference. Fault tolerance is thus one of the key requirements for data communications in WSN applications. This paper proposes a novel path redundancy-based algorithm, called dual separate paths (DSP), that provides fault-tolerant communication with the improvement of the network traffic performance for WSN applications, such as fault-tolerant CPSs. The proposed DSP algorithm establishes two separate paths between a source and a destination in a network based on the network topology information. These paths are node-disjoint paths and have optimal path distances. Unicast frames are delivered from the source to the destination in the network through the dual paths, providing fault-tolerant communication and reducing redundant unicast traffic for the network. The DSP algorithm can be applied to wired and wireless networks, such as WSNs, to provide seamless fault-tolerant communication for mission-critical and life-critical applications such as fault-tolerant CPSs. The analyzed and simulated results show that the DSP-based approach not only provides fault-tolerant communication, but also improves network traffic performance. For the case study in this paper, when the DSP algorithm was applied to high-availability seamless redundancy (HSR) networks, the proposed DSP-based approach reduced the network traffic by 80% to 88% compared with the standard HSR protocol, thus improving network traffic performance.