Safety-related Applications Research Articles

New generation steels for light weight vehicle safety related applications

Abstract Vehicle lightening efforts, low fuel consumption and low CO2 emission have become an important target in the automotive industry without compromising safety and comfort. By reducing vehicle weights, fuel consumption and CO2 emissions are significantly reduced. In this context, the next generation steels are constantly being developed in terms of high strength, forming, weld ability, low cost in matters such as vehicle lightening and impact safety. In this study, side-impact performance of the next generation steels that can be used for the body B-Pillar part as part of vehicle lightening was examined. In finite element analysis, Usibor 1500, TWIP 980, TRIP 800, DP 1000 materials and three different metal sheet thicknesses (1.3–1.5–1.8 mm) were used for the B-Pillar part. All results obtained were compared in terms of energy absorption ability, formability, weld ability, cost and weight.

Studies on thermodynamics properties of binary mixtures of amyl acetate with propanol at T = 298.15 K to 318.15 K

The thermophysical properties of amyl acetate and isomers of propanol are critical for a variety of industrial, scientific, and safety–related applications. The density (ρ), viscosity (η), speed of sound (u), and refractive index (nD) of binary amyl acetate (1) + propanol (2) mixtures were measured at T = 298.15 K–318.15 K and pressure of 0.1 MPa. The excess molar volume (VmE), deviation in viscosity (Δη), deviation in speed of sound (Δu), excess isentropic compressibility (κsE), and deviation in refractive index (ΔnD) were derived from measured data. The deviation parameters in binary mixtures are affected by H–bonding, interstitial accommodation, and dipole–dipole interactions. The VmE, Δη and κsE values were described with Graph theory, Prigogine-Flory-Patterson (PFP) theory, Singh model, and ab–initio approach. While the VmE and Δη rise with increment in temperature whereas Δu and ΔnD decrease with increase in temperature.

Acoustic emission of lattice structures under cycling loading relates process parameters with fatigue properties

Metamaterials, especially lattice structures, are of great interest for many application areas such as aerospace, automotive and medicine due to their adjustable mechanical properties and their low weight. Due to their complex geometry, lattice structures are usually manufactured additively, which causes a large variance in the manufacturing-related mechanical properties. In order to establish metamaterials in industrial applications under cyclic loading, the fatigue behavior needs to be investigated to evaluate the load capacity of these structures. Here we analyze the fatigue behavior of AlSi10Mg truss structures fabricated with L-PBF using a load increase test in combination with acoustic emission measurements. The acoustic signals are evaluated in terms of time-dependent amplitude signal and frequency spectrum. Increasing load and increasing specimen damage resulted in changes of the acoustic spectrum and the amplitude of the time signal. Based on the results, a correlation of specimen properties with build platform position in the manufacturing process could be established. Acoustic emission measurement as an in situ characterization method during cyclic loading is promising for surveillance of lattice structures in safety related applications.

Age of Information, Latency, and Reliability in Intelligent Vehicular Networks

Autonomous driving and intelligent transportation have posed stringent requirements on different performance metrics of vehicle-to-everything (V2X) communications depending on the wireless service being carried. Such coexisting heterogeneous performance requirements bring about increasing challenge in managing modern vehicular networks. In this article, we categorize wireless vehicular traffic into two kinds, namely, status update service and data communication service, where in safety-related applications the former is concerned with age of information (AoI) while the latter cares about latency and reliability in general. Afterwards, we uncover the relationship among AoI, latency, and reliability, and propose a novel graphical packet delivery model that is powerful for analyzing AoI performance. To unify the AoI, latency, and reliability management for different services, we further develop an intelligent resource allocation framework validated by a quantitative case study. Finally, we explore key challenges and opportunities to address the coexistence of status update and data communication services in intelligent vehicular networks.

An RSU-Assisted Hybrid Emergency Message Broadcasting Protocol for VANETs

In vehicular ad hoc networks (VANETs), emergency messages broadcasting has been considered as one of the important parts for safety related applications. The efficiency and reliability of the emergency message broadcasting are severely affected by the network structure, message redundancy, channel contention, etc. To address this problem, in this paper, we propose an RSUs-assisted hybrid emergency message broadcasting (RA-HEMB) protocol for two-way grid roads in urban VANETs. In doing so, the broadcasting target region is first generated based on the instantaneous traffic status and types of the emergency messages. To balance the deliver latency and the reliability, an adaptive forwarding nodes selection scheme based on position is proposed to dynamically determine the forwarding area according to different vehicle densities and vehicle-to-vehicle (V2V) communication ranges. Then, based on the forwarding nodes selection scheme and roadside unit (RSU) deployment, an RA-HEMB protocol is established by properly utilizing both the wired links via RSUs and the wireless links to broadcast the emergency message in hybrid vehicular networks. Simulation results show that the proposed RA-HEMB can greatly shorten the broadcasting time needed to notice the nodes within the target region performance compared to the benchmark protocols.

Adversarial Robust Aerial Image Recognition Based on Reactive-Proactive Defense Framework with Deep Ensembles

As a safety-related application, visual systems based on deep neural networks (DNNs) in modern unmanned aerial vehicles (UAVs) show adversarial vulnerability when performing real-time inference. Recently, deep ensembles with various defensive strategies against adversarial samples have drawn much attention due to the increased diversity and reduced variance for their members. Aimed at the recognition task of remote sensing images (RSIs), this paper proposes to use a reactive-proactive ensemble defense framework to solve the security problem. In reactive defense, we fuse scoring functions of several classical detection algorithms with the hidden features and average output confidences from sub-models as a second fusion. In terms of proactive defense, we attempt two strategies, including enhancing the robustness of each sub-model and limiting the transferability among sub-models. In practical applications, the real-time RSIs are first input to the reactive defense part, which can detect and reject the adversarial RSIs. The accepted ones are then passed to robust recognition with a proactive defense. We conduct extensive experiments on three benchmark RSI datasets (i.e., UCM, AID, and FGSC-23). The experimental results show that the deep ensemble method of reactive and proactive defense performs very well in gradient-based attacks. The analysis of the applicable attack scenarios for each proactive ensemble defense is also helpful for this field. We also perform a case study with the whole framework in the black-box scenario, and the highest detection rate reaches 93.25%. Most of the adversarial RSIs can be rejected in advance or correctly recognized by the enhanced deep ensemble. This article is the first one to combine reactive and proactive defenses with a deep ensemble against adversarial attacks in the context of RSI recognition for DNN-based UAVs.

Assay data of spent nuclear fuel: the lab-work behind the numbers

Computational modelling for spent nuclear fuel (SNF) characterization is already widely used and continuously further developed for a plethora of safety related applications and licensing issues in SNF management. An essential step in the development of these methodologies is the validation: the demonstration that the SNF elemental and isotopic composition is sufficiently accurately predicted by the code calculations. This validation step requires experimentally measured nuclide concentrations in SNF, together with an estimation of related uncertainties. The SFCOMPO 2.0 database of OECD/NEA is a database of such publicly available assay data of SNF. A basic understanding of all analytical steps that finally result in assay data of SNF is important for modelers when assessing the “fit-for-validation” requirement of an experimental dataset. The aim of this article is to explain users of such datasets the complex analytical pathway towards assay data. Points of attention, challenges and pitfalls all along the analytical pathway will be discussed, from sampling, dissolution procedures, necessary dilutions and separations, available analytical techniques, some related uncertainties, up to reporting of the results.

Artificial Intelligence and Deep Learning-Based System for Agri-Food Quality and Safety Detection

Deep Learning (DL) has emerged as a highly effective technique for analyzing large volumes of data across various domains, including image processing, speech recognition, and pattern recognition. Recently, DL has also found applications in the field of food science and engineering, a relatively novel area of research. This paper provides a concise introduction to DL and delves into the architecture of a typical Convolution Neural Network (CNN) structure, as well as AI and IoT (Internet of Things) data training methodologies. Our research involved an extensive review of studies that utilized DL as a computational approach to address food-related challenges, such as food recognition, calorie computation, and safety detection of various food types like fruits, potatoes, meats, and aquatic products, as well as food supply chain management and food borne illness detection. Each study examined different problems, datasets, preprocessing techniques, network architectures, and evaluation metrics, comparing their results with alternative solutions. Furthermore, we explored the role of big data in the field of food quality assurance, uncovering compelling trends. Based on our analysis, DL consistently outperforms other approaches, including manual feature extractors and traditional machine learning algorithms. The findings highlight the tremendous potential of DL as a promising technology for food safety inspections and related applications in the food industry

Multispectral Invisible Coating: Laminated Visible-Thermal Physical Attack against Multispectral Object Detectors Using Transparent Low-E Films

Multispectral object detection plays a vital role in safety-critical vision systems that require an around-the-clock operation and encounter dynamic real-world situations(e.g., self-driving cars and autonomous surveillance systems). Despite its crucial competence in safety-related applications, its security against physical attacks is severely understudied. We investigate the vulnerability of multispectral detectors against physical attacks by proposing a new physical method: Multispectral Invisible Coating. Utilizing transparent Low-e films, we realize a laminated visible-thermal physical attack by attaching Low-e films over a visible attack printing. Moreover, we apply our physical method to manufacture a Multispectral Invisible Suit that hides persons from the multiple view angles of Multispectral detectors. To simulate our attack under various surveillance scenes, we constructed a large-scale multispectral pedestrian dataset which we will release in public. Extensive experiments show that our proposed method effectively attacks the state-of-the-art multispectral detector both in the digital space and the physical world.

Adversarial Robustness Enhancement of UAV-Oriented Automatic Image Recognition Based on Deep Ensemble Models

Deep neural networks (DNNs) have been widely utilized in automatic visual navigation and recognition on modern unmanned aerial vehicles (UAVs), achieving state-of-the-art performances. However, DNN-based visual recognition systems on UAVs show serious vulnerability to adversarial camouflage patterns on targets and well-designed imperceptible perturbations in real-time images, which poses a threat to safety-related applications. Considering a scenario in which a UAV is suffering from adversarial attack, in this paper, we investigate and construct two ensemble approaches with CNN and transformer for both proactive (i.e., generate robust models) and reactive (i.e., adversarial detection) adversarial defense. They are expected to be secure under attack and adapt to the resource-limited environment on UAVs. Specifically, the probability distributions of output layers from base DNN models in the ensemble are combined in the proactive defense, which mainly exploits the weak adversarial transferability between the CNN and transformer. For the reactive defense, we integrate the scoring functions of several adversarial detectors with the hidden features and average the output confidence scores from ResNets and ViTs as a second integration. To verify their effectiveness in the recognition task of remote sensing images, we conduct experiments on both optical and synthetic aperture radar (SAR) datasets. We find that the ensemble model in proactive defense performs as well as three popular counterparts, and both of the ensemble approaches can achieve much more satisfactory results than a single base model/detector, which effectively alleviates adversarial vulnerability without extra re-training. In addition, we establish a one-stop platform for conveniently evaluating adversarial robustness and performing defense on recognition models called AREP-RSIs, which is beneficial for the future research of the remote sensing field.

Target set selection in social networks with tiered influence and activation thresholds.

Thanks to the mass adoption of internet and mobile devices, users of the social media can seamlessly and spontaneously connect with their friends, followers and followees. Consequently, social media networks have gradually become the major venue for broadcasting and relaying information, and is casting great influences on the people in many aspects of their daily lives. Thus locating those influential users in social media has become crucially important for the successes of many viral marketing, cyber security, politics, and safety-related applications. In this study, we address the problem through solving the tiered influence and activation thresholds target set selection problem, which is to find the seed nodes that can influence the most users within a limited time frame. Both the minimum influential seeds and maximum influence within budget problems are considered in this study. Besides, this study proposes several models exploiting different requirements on seed nodes selection, such as maximum activation, early activation and dynamic threshold. These time-indexed integer program models suffer from the computational difficulties due to the large numbers of binary variables to model influence actions at each time epoch. To address this challenge, this paper designs and leverages several efficient algorithms, i.e., Graph Partition, Nodes Selection, Greedy algorithm, recursive threshold back algorithm and two-stage approach in time, especially for large-scale networks. Computational results show that it is beneficial to apply either the breadth first search or depth first search greedy algorithms for the large instances. In addition, algorithms based on node selection methods perform better in the long-tailed networks.

FRSR: Framework for real-time scene reconstruction in robot-assisted minimally invasive surgery

3D reconstruction of the intra-operative scenes provides precise position information which is the foundation of various safety related applications in robot-assisted surgery, such as augmented reality. Herein, a framework integrated into a known surgical system is proposed to enhance the safety of robotic surgery. In this paper, we present a scene reconstruction framework to restore the 3D information of the surgical site in real time. In particular, a lightweight encoder–decoder network is designed to perform disparity estimation, which is the key component of the scene reconstruction framework. The stereo endoscope of da Vinci Research Kit (dVRK) is adopted to explore the feasibility of the proposed approach, and it provides the possibility for the migration to other Robot Operating System (ROS) based robot platforms due to the strong independence on hardware. The framework is evaluated using three different scenarios, including a public dataset (3018 pairs of endoscopic images), the scene from the dVRK endoscope in our lab as well as a self-made clinical dataset captured from an oncology hospital. Experimental results show that the proposed framework can reconstruct 3D surgical scenes in real time (25 FPS), and achieve high accuracy (2.69 ± 1.48 mm in MAE, 5.47 ± 1.34 mm in RMSE and 0.41 ± 0.23 in SRE, respectively). It demonstrates that our framework can reconstruct intra-operative scenes with high reliability of both accuracy and speed, and the validation of clinical data also shows its potential in surgery. This work enhances the state of art in 3D intra-operative scene reconstruction based on medical robot platforms. The clinical dataset has been released to promote the development of scene reconstruction in the medical image community.

Integrity monitoring scheme for single-epoch GNSS PPP-RTK positioning

Integrity monitoring for precise point positioning is critical for safety-related applications. With the increasing demands of high-accuracy autonomous navigation for unmanned ground and aerial vehicles, the integrity monitoring method of high-precision positioning has become an essential requirement. While high precision Global Navigation Satellite Systems (GNSS) positioning is widely used in such applications, there are still many difficulties in the integrity monitoring method for the multi-frequency multi-GNSS undifferenced and uncombined Precise Point Positioning (PPP). The main difficulties are caused by using the measurements of multiple epochs in PPP. Based on the baseline Multiple Hypothesis Solution Separation (MHSS) Advanced Receiver Autonomous Integrity Monitoring (ARAIM) algorithm, this paper discusses the feasibility of the pseudorange-based baseline ARAIM method on the single-epoch PPP based on Real-Time Kinematic (RTK) networks (PPP-RTK) framework to overcome these difficulties. In addition, a new scheme is proposed to transfer the conventional PPP process into the single-epoch PPP-RTK framework. The simulation results using the proposed model are analyzed in this study. The Protection Levels (PLs) estimated by PPP Wide-lane Ambiguity Resolution (PPP-WAR) model with regional corrections can reach the meter level and the PLs estimated by PPP Ambiguity Resolution (PPP-AR) and PPP-RTK models are usually the sub-meter level. Given a horizontal Alert Limit (AL) of 1.5 m, the global coverage of availability above 99.9% for PPP-WAR, PPP-AR, and PPP-RTK can reach 92.6%, 99.4%, and 99.7% respectively. The results using real kinematic data also show that tight PLs can be achieved when the observation conditions are good.

A Fast, Reliable, Opportunistic Broadcast Scheme With Mitigation of Internal Interference in VANETs

In VANETs, it is important to support fast and reliable multi-hop broadcast for safety-related applications. The performance of multi-hop broadcast schemes is greatly affected by relay selection strategies. However, the relationship between the relay selection strategies and the expected broadcast performance has not been fully characterized yet. Furthermore, conventional broadcast schemes usually attempt to minimize the waiting time difference between adjacent relay candidates to reduce the waiting time overhead, which makes the relay selection process vulnerable to internal interference, occurring due to retransmissions from previous forwarders and transmissions from redundant relays. In this paper, we jointly take both of the relay selection and the internal interference mitigation into account and propose a fast, reliable, opportunistic multi-hop broadcast scheme, in which we utilize a novel metric called the expected broadcast speed in relay selection and propose a delayed retransmission mechanism to mitigate the adverse effect of retransmissions from previous forwarders and an expected redundancy probability based mechanism to mitigate the adverse effect of redundant relays. The performance evaluation results show that the proposed scheme yields the best broadcast performance among the four schemes in terms of the broadcast coverage ratio and the end-to-end delivery latency.

Delay-Tolerant Distributed Algorithms for Decision-Making in Vehicular Networks

Learning a fast global model that describes the observed phenomenon well is a crucial goal in the inherently distributed Vehicular Networks. This global model is further used for decision-making, which is especially important for some safety-related applications (i.e., the altering of accident and warning of traffic jam). Most existing works have ignored the network overhead caused by synchronizing with neighbors, which inevitably delays the time for agents to stabilize. In this paper, we focus on developing an asynchronous distributed clustering algorithm to learn the global model, where cluster models, rather than raw data points, are shared and updated. Empirical experiments on a message delay simulator show the efficiency of our methods, with a reduced convergence time, declined network overhead and improved accuracy (relative to the standard solution). This algorithm is further improved by introducing a tolerant delay. Compared to the algorithm without delay, the performance is improved significantly in terms of convergence time (by as much as 47%) and network overhead (by around 53%) if the underlying network is geometric or regular.

Design and Suitability Analysis of FPGA-Based TCP/IP Offload Module for Control Systems of FBRs

Transmission Control Protocol/Internet Protocol (TCP/IP) offload modules play a crucial role in India’s first commercial fast breeder reactor (PFBR) distributed digital control system architecture by providing networking capabilities to the real-time control systems used in safety-critical and safety-related applications of the PFBR. These modules enable operators to remotely control various mechanisms, facilities, and process loops via real-time control systems from the main control room display stations. The malfunctioning of these modules may hamper the overall plant’s availability; hence to improve further upon safety, security, reliability, and performance, these modules have been indigenously developed for upcoming fast breeder reactors in India. This paper proposes methods to justify the suitability of these communication modules as per the D-25 design guidelines issued by the Atomic Energy Regulatory Board of India. The design and testing aspects are covered, including an approach toward hardware descriptive language-based design and testing with respect to the IEEE standard and the TCP/IP protocol specifications. Module design validation based on ethernet frame transactions is detailed in this paper. The test platform design, which includes the diagnostics software design for CPUs designed with these TCP offload modules, and the design of a graphical user interface to detect and record TCP connection anomalies are detailed. The network interface response to postulated events of TCP packet corruption, packet drop, packet delay, out-of-order packets, duplicate packets, and heavy traffic is characterized.

Detection and Reporting of Wireless Channel Congestion and Interference in Connected Vehicle Networks

Connected vehicle (CV) wireless networks based on dedicated short-range communications (DSRC), European Telecommunications Standards Institute (ETSI) intelligent transportation systems (ITS-G5), and C-V2X technologies are susceptible to interference from both unintentional emitters and non-CV devices that may be authorized to share the same or adjacent bands. Such interference may lead to unreliable communication and disruption of CV services with a particular impact on safety-related applications. Surprisingly, considering the safety-critical nature of CV applications, there is no simple mechanism for detecting congestion or interference in such networks over wide areas. To address this gap, we propose and demonstrate that both interference and congestion in DSRC and ETSI ITS-G5 networks can be detected simply and inexpensively using capabilities that are already incorporated into the IEEE 802.11p standard, specifically the flags and statistics generated mostly in the physical layer (physical layer convergence procedure and physical medium dependent) state machines. Such a capability could be realized through a relatively minor software upgrade but would resolve a longstanding but underappreciated concern that CV networks are vulnerable to both congestion and a variety of short-range interferers but lack the capability to detect or report this. Although our focus was on DSRC and ITS-G5, similar considerations apply to related schemes such as C-V2X.

Dispersion modeling of underwater oil released from buried subsea pipeline considering current and wave

Marine oil release accident may pose a serve threat to marine environment. Modeling of underwater oil dispersion is helpful for risk assessment and emergency response of marine oil spill accident. This paper presents a Computational Fluid Dynamics (CFD) model to describe the dispersion behavior of oil released from buried subsea oil pipeline leak considering current and wave. The numerical model is validated by repeating the small-scale experiments. A comprehensive simulation of underwater oil dispersion from seafloor soil to sea surface under the combined action of wave and current is carried out. Effects of oil density, oil release rate, soil porosity, current speed on underwater oil dispersion are analyzed. Besides, this study considers underwater oil dispersion from multiple leak points on the pipeline, and a regression model for rapidly predicting oil dispersion is developed by numerical fitting. The uniqueness of this study is that it models underwater oil dispersion considering the seafloor soil and the combined action of current and wave. The presence of seafloor soil has a significant effect on oil migration shape and trajectory, and the variation of soil porosity has little effect on the dispersion distance of crude oil plume. Essentially, this study can be useful for the safety-related applications in marine oil spill accidents.

Load Estimation of Moving Passenger Cars Using Inductive-Loop Technology.

Due to their lack of driving controllability, overweight vehicles are a big threat to road safety. The proposed method for a moving passenger car load estimation is capable of detecting an overweight vehicle, and thus it finds its application in road safety improvement. The weight of a car's load entering or leaving a considered zone, e.g., industrial facility, a state, etc., is also of concern in many applications, e.g., surveillance. Dedicated vehicle weight-in-motion measurement systems generally use expensive load sensors that also require deep intervention in the road while being installed and also are calibrated only for heavy trucks. In this paper, a vehicle magnetic profile (VMP) is used for defining a load parameter proportional to the passenger vehicle load. The usefulness of the proposed load parameter is experimentally demonstrated in field tests. The sensitivity of the VMP to the load change results from the fact that the higher load decreases the vehicle clearance value which in turn increases the VMP. It is also shown that a slim inductive-loop sensors allows the building of a load estimation system, with a maximum error around 30 kg, which allows approximate determination of the number of passengers in the car. The presented proof of concept extends the functionality of inductive loops, already installed in the road, for acquiring other traffic parameters, e.g., moving vehicle axle-to-axle distance measurement, to road safety and surveillance related applications.

How I Met Your V2X Sensor Data: Analysis of Projection-Based Light Field Visualization for Vehicle-to-Everything Communication Protocols and Use Cases.

The practical usage of V2X communication protocols started emerging in recent years. Data built on sensor information are displayed via onboard units and smart devices. However, perceptually obtaining such data may be counterproductive in terms of visual attention, particularly in the case of safety-related applications. Using the windshield as a display may solve this issue, but switching between 2D information and the 3D reality of traffic may introduce issues of its own. To overcome such difficulties, automotive light field visualization is introduced. In this paper, we investigate the visualization of V2X communication protocols and use cases via projection-based light field technology. Our work is motivated by the abundance of V2X sensor data, the low latency of V2X data transfer, the availability of automotive light field prototypes, the prevalent dominance of non-autonomous and non-remote driving, and the lack of V2X-based light field solutions. As our primary contributions, we provide a comprehensive technological review of light field and V2X communication, a set of recommendations for design and implementation, an extensive discussion and implication analysis, the exploration of utilization based on standardized protocols, and use-case-specific considerations.