Showing papers on "System safety published in 2019"

Ensemble machine learning models for aviation incident risk prediction

Abstract: With the spectacular growth of air traffic demand expected over the next two decades, the safety of the air transportation system is of increasing concern. In this paper, we facilitate the “proactive safety” paradigm to increase system safety with a focus on predicting the severity of abnormal aviation events in terms of their risk levels. To accomplish this goal, a predictive model needs to be developed to examine a wide variety of possible cases and quantify the risk associated with the possible outcome. By utilizing the incident reports available in the Aviation Safety Reporting System (ASRS), we build a hybrid model consisting of support vector machine and an ensemble of deep neural networks to quantify the risk associated with the consequence of each hazardous cause. The proposed methodology is developed in four steps. First, we categorize all the events, based on the level of risk associated with the event consequence, into five groups: high risk, moderately high risk, medium risk, moderately medium risk, and low risk. Secondly, a support vector machine model is used to discover the relationships between the event synopsis in text format and event consequence. In parallel, an ensemble of deep neural networks is trained to model the intricate associations between event contextual features and event outcomes. Thirdly, an innovative fusion rule is developed to blend the prediction results from the two types of trained machine learning models, thereby improving the prediction. Finally, the prediction on risk level categorization is extended to event-level outcomes through a probabilistic decision tree. By comparing the performance of the developed hybrid model against another three individual models with ten-fold cross-validation and statistical tests, we demonstrate the effectiveness of hybrid model in quantifying the risk related to the consequences of hazardous events.

A BIM approach for construction safety: applications, barriers and solutions

Abstract: Construction industry bears a lot of casualties and accidents more than other high-risk industries annually. Thus, the use of new technologies such, as building information modeling, automatic rule checking, information technology-based safety systems in order to implement the rules and safety standards, better controls the performance of workers on site and make high coordination between operational executives, leading to create a secure environment in projects by reducing accidents. The paper aims to discuss these issues.,In this study, a researcher-designed questionnaire was distributed among 200 companies that are active in the field of construction to evaluate the effect of building information model (BIM) for safety projects and barriers to adoption. Only 70 percent of questionnaires were returned. Statistical Package for the Social Sciences analysis has been used to determine the correlation coefficient among the respondents.,The results show the factors that lead to failure in the adoption of BIM in Iran are lack of well-trained personnel, proper social infrastructure, guidance and governmental supports.,Finally, the authors presented solutions for overcoming barriers and proposed some factors leading to the successful adoption of BIM in Iran.

Towards a Framework for Safety Assurance of Autonomous Systems

Abstract: Autonomous systems have the potential to provide great benefit to society. However, they also pose problems for safety assurance, whether fully auton-omous or remotely operated (semi-autonomous). This paper discusses the challenges of safety assur-ance of autonomous systems and proposes a novel framework for safety assurance that, inter alia, uses machine learning to provide evidence for a system safety case and thus enables the safety case to be updated dynamically as system behaviour evolves.

Trends in vehicle motion control for automated driving on public roads

Abstract: In this paper, we describe how vehicle systems and the vehicle motion control are affected by automated driving on public roads. We describe the redundancy needed for a road vehicle to meet certain safety goals. The concept of system safety as well as system solutions to fault tolerant actuation of steering and braking and the associated fault tolerant power supply is described. Notably restriction of the operational domain in case of reduced capability of the driving automation system is discussed. Further we consider path tracking, state estimation of vehicle motion control required for automated driving as well as an example of a minimum risk manoeuver and redundant steering by means of differential braking. The steering by differential braking could offer heterogeneous or dissimilar redundancy that complements the redundancy of described fault tolerant steering systems for driving automation equipped vehicles. Finally, the important topic of verification of driving automation systems is addressed.

On the Safe Road Toward Autonomous Driving: Phase noise monitoring in radar sensors for functional safety compliance

Abstract: The first approaches to improve vehicle safety were so-called passive safety systems, which did not directly interfere with the driving process but protected the occupants during a crash. In contrast, the first assistance system was the antilock braking system (ABS) successfully introduced in the early 1970s. This active system was developed to avoid an accident by automatically intervening in the braking behavior of the car. At about the same time, the first automotive radar prototype was presented. Since the invention of this very unwieldy radar system, organizations all around the world spent significant efforts in pushing the development of automotive radar systems forward. Today, radar sensors together with ultrasound sensors, lidar, and cameras form the backbone of advanced driver assistant systems (ADASs) as well as autonomous driving (AD), which is in the prototype stage. In particular, because of their robustness against adverse lighting and weather conditions, radar sensors are considered a key technology for modern vehicle safety and comfort systems. Along with the trend toward higher automation, more cars will be equipped with radar sensors in the near future. Because ADASs directly influence the vehicle dynamics, new regulating functional safety (FuSa) requirements, such as the ISO 26262 standard, were introduced. These requirements are mandatory to protect the road users.

Integrating Model Checking With SysML in Complex System Safety Analysis

Abstract: Modern complex systems are characterized by numerous complex interactions and high levels of integration of functions, which present new challenges from the viewpoints of system safety analysis and design. Model checking can be employed to perform safety analysis, identify potential hazards, and prove the correctness of complex systems. However, many types of construction models are expressed in different ways, and there exists no unified model. Thus, the integration of model checking with system modeling language is proposed herein to analyze the safety of complex systems. System modeling language (SysML) is introduced to establish a unified system model that can describe a hybrid system of hardware and software but cannot be applied directly to safety analysis. Therefore, the semi-formal model SysML is transformed into the formal model new symbolic model checker/verifier, and the transformation rules are defined. The proposed unified model can not only help designers and safety and software engineers to execute various tasks but also efficiently, completely, and accurately analyze and verify the safety of complex systems. Finally, an integrated modular avionics case is presented to illustrate how to analyze the safety of complex systems. The results of the case study show that the proposed method can help increase the efficiency of safety analysis work and improve system safety.

Impacts of Lean Construction on Safety Systems: A System Dynamics Approach.

Abstract: Lean construction has been viewed as an effective management approach for reducing the occurrence of no-value or destructive activities, such as wasting resources and safety-related accidents. However, few studies have systematically addressed how and to what extent lean construction practices influence construction safety. To bridge this gap, a conceptual model is developed and validated using a system dynamics approach. The construction system in this model comprises four sub-systems (i.e., environment system, equipment system, management system, and employee system). Data were collected from 448 projects in China. Simulations were conducted to determine the correlations between five types of lean tools and the four construction sub-systems. The results show that: (a) 5S management has significant positive impacts on the control of key locations and facilities at construction sites, and contributes to the mitigation of environmental impacts; (b) visual management can significantly improve safety compliance and safety management; (c) just-in-time management has significantly positive influences on the safety facilities layout and formulation of the safety plan; and (d) the Last Planner® System and conference management are effective in improving safety training and the implementation of the safety plan. These findings provide new insights into the use of lean construction for improving construction safety through the implementation of a targeted lean approach.

RBD Model-Based Approach for Reliability Assessment in Complex Systems

Abstract: Nowadays, system reliability performance represents a key issue in any advanced technology application in order to guarantee ambient, personnel, and system safety. The core of this paper is the reliability block diagram (RBD) generation with the aim of providing project engineers a reliability prediction in the early stages of industrial product development. This study is focused on gas turbine auxiliary systems; these systems include both mechanical items (in particular hydraulic devices such as valves, pumps, and filters) and electronic ones (e.g., sensors, instruments, control logic). Such complex structures are decomposed in single blocks and interconnections to establish their mutual relationship and achieve reliability performance of the whole system. The case study is one of the most important gas turbine auxiliary systems, the mineral lube oil console. The aim of this paper is to introduce a new approach to assess system reliability prediction in presence of redundant and stand-by architectures; redundancy is widely used in industrial applications since it is one of the best techniques to achieve fault tolerance. The proposed method led to the development of a dedicated software tool RBDesigner that semiautomatically generates a RBD starting from the sketch of thermal-hydraulic systems and provides the most important reliability parameters. The use of the proposed tool allows project engineers to reduce time delivery, reduce time for improvements, achieve reliability targets, and guarantee availability performance to the customers. The strengths of RBDesigner were finally validated by a comparison with other commercial software solutions.

Adaptive Safety Context Information for Vulnerable Road Users with MEC Support

Abstract: Cooperative Vulnerable Road User (VRU) collision avoidance aims at preventing potential accidents between VRUs and vehicles by exchanging context information. In this paper, we present a Multi-access Edge Computing (MEC)-based VRU safety system as an alternative to earlier purely ad-hoc communication-based ones, in which VRU smartphones utilize the cellular connection to frequently send context messages to a MEC server. However, in such safety systems, calculating context information on smartphones, which are already resource-restricted, could lead to reduced battery lifetime and, thus, to poor user experiences. To deal with this issue, we propose an adaptive approach for VRU context information calculation, which considers the use of computation offloading when needed in order to save energy while still ensuring timeliness. As a baseline, we use our machine learning application for determining pedestrian activities. Both experimental and simulation results suggest that it is worth to offload context information computation to the MEC when the updating interval or the sensor sampling frequency is low, i.e., the amount of raw data collected is small; otherwise, local execution is preferable. We see our results as a basis for designing more energy-efficiency calculation models for VRU safety systems.

Decentralized Real-Time Safety Verification for Distributed Cyber-Physical Systems

Abstract: Safety-critical distributed cyber-physical systems (CPSs) have been found in a wide range of applications. Notably, they have displayed a great deal of utility in intelligent transportation, where autonomous vehicles communicate and cooperate with each other via a high-speed communication network. Such systems require an ability to identify maneuvers in real-time that cause dangerous circumstances and ensure the implementation always meets safety-critical requirements. In this paper, we propose a real-time decentralized safety verification approach for a distributed multi-agent CPS with the underlying assumption that all agents are time-synchronized with a low degree of error. In the proposed approach, each agent periodically computes its local reachable set and exchanges this reachable set with the other agents with the goal of verifying the system safety. Our method, implemented in Java, takes advantages of the timing information and the reachable set information that are available in the exchanged messages to reason about the safety of the whole system in a decentralized manner. Any particular agent can also perform local safety verification tasks based on their local clocks by analyzing the messages it receives. We applied the proposed method to verify, in real-time, the safety properties of a group of quadcopters performing a distributed search mission.

Highly parallelized data-driven MPC for minimal intervention shared control

Abstract: We present a shared control paradigm that improves a user's ability to operate complex, dynamic systems in potentially dangerous environments without a priori knowledge of the user's objective. In this paradigm, the role of the autonomous partner is to improve the general safety of the system without constraining the user's ability to achieve unspecified behaviors. Our approach relies on a data-driven, model-based representation of the joint human-machine system to evaluate, in parallel, a significant number of potential inputs that the user may wish to provide. These samples are used to (1) predict the safety of the system over a receding horizon, and (2) minimize the influence of the autonomous partner. The resulting shared control algorithm maximizes the authority allocated to the human partner to improve their sense of agency, while improving safety. We evaluate the efficacy of our shared control algorithm with a human subjects study (n=20) conducted in two simulated environments: a balance bot and a race car. During the experiment, users are free to operate each system however they would like (i.e., there is no specified task) and are only asked to try to avoid unsafe regions of the state space. Using modern computational resources (i.e., GPUs) our approach is able to consider more than 10,000 potential trajectories at each time step in a control loop running at 100Hz for the balance bot and 60Hz for the race car. The results of the study show that our shared control paradigm improves system safety without knowledge of the user's goal, while maintaining high-levels of user satisfaction and low-levels of frustration. Our code is available online at this https URL.

Hazard Analysis for Escalator Emergency Braking System via System Safety Analysis Method Based on STAMP

Abstract: Due to the complex mechanical structure and control process of escalator emergency braking systems (EEBS), traditional hazard analysis based on the event chain model have limitations in exploring component interaction failure in such a complex social-technical system. Therefore, a hazard analysis framework is proposed in this paper for hazard analysis of complex electromechanical systems based on system-theoretic accident model and process (STAMP). Firstly, basic principles of STAMP are introduced and comparison with other hazard analysis methods is conducted, then the safety analysis framework is proposed. Secondly, a study case is performed to identify unsafe control actions of EEBS from control structures, and a specific control diagram is organized to recognize potential example casual scenarios. Next, comparison between fault tree analysis and STAMP for escalator’s overturned accident shows that hazards related to component damaged can be identified by both, while hazards that focus on components interaction can only be identified by STAMP. Besides, single control way and tandem operation process are found to be the obvious causal factors of accidents. Finally, some improvement measures like decibel detection or vibration monitoring of key components are suggested to help the current broken chain detection to trigger the anti-reversal device for a better safe EEBS.

Learning-Based Time Delay Attack Characterization for Cyber-Physical Systems

Abstract: The cyber-physical systems (CPSes) rely on computing and control techniques to achieve system safety and reliability. However, recent attacks show that these techniques are vulnerable once the cyber-attackers have bypassed air gaps. The attacks may cause service disruptions or even physical damages. This paper designs the built-in attack characterization scheme for one general type of cyber-attacks in CPS, which we call time delay attack, that delays the transmission of the system control commands. We use the recurrent neural networks in deep learning to estimate the delay values from the input trace. Specifically, to deal with the long time-sequence data, we design the deep learning model using stacked bidirectional long short-term memory (LSTM) units. The proposed approach is tested by using the data generated from a power plant control system. The results show that the LSTM-based deep learning approach can work well based on data traces from three sensor measurements, i.e., temperature, pressure, and power generation, in the power plant control system. Moreover, we show that the proposed approach outperforms the base approach based on k-nearest neighbors.

Leveraging artifact trees to evolve and reuse safety cases

Abstract: Safety Assurance Cases (SACs) are increasingly used to guide and evaluate the safety of software-intensive systems. They are used to construct a hierarchically organized set of claims, arguments, and evidence in order to provide a structured argument that a system is safe for use. However, as the system evolves and grows in size, a SAC can be difficult to maintain. In this paper we utilize design science to develop a novel solution for identifying areas of a SAC that are affected by changes to the system. Moreover, we generate actionable recommendations for updating the SAC, including its underlying artifacts and trace links, in order to evolve an existing safety case for use in a new version of the system. Our approach, Safety Artifact Forest Analysis (SAFA), leverages traceability to automatically compare software artifacts from a previously approved or certified version with a new version of the system. We identify, visualize, and explain changes in a Delta Tree. We evaluate our approach using the Dronology system for monitoring and coordinating the actions of cooperating, small Unmanned Aerial Vehicles. Results from a user study show that SAFA helped users to identify changes that potentially impacted system safety and provided information that could be used to help maintain and evolve a SAC1.

Highly Parallelized Data-driven MPC for Minimal Intervention Shared Control

Abstract: We present a shared control paradigm that improves a user's ability to operate complex, dynamic systems in potentially dangerous environments without a priori knowledge of the user's objective. In this paradigm, the role of the autonomous partner is to improve the general safety of the system without constraining the user's ability to achieve unspecified behaviors. Our approach relies on a data-driven, model-based representation of the joint human-machine system to evaluate, in parallel, a significant number of potential inputs that the user may wish to provide. These samples are used to (1) predict the safety of the system over a receding horizon, and (2) minimize the influence of the autonomous partner. The resulting shared control algorithm maximizes the authority allocated to the human partner to improve their sense of agency, while improving safety. We evaluate the efficacy of our shared control algorithm with a human subjects study (n=20) conducted in two simulated environments: a balance bot and a race car. During the experiment, users are free to operate each system however they would like (i.e., there is no specified task) and are only asked to try to avoid unsafe regions of the state space. Using modern computational resources (i.e., GPUs) our approach is able to consider more than 10,000 potential trajectories at each time step in a control loop running at 100Hz for the balance bot and 60Hz for the race car. The results of the study show that our shared control paradigm improves system safety without knowledge of the user's goal, while maintaining high-levels of user satisfaction and low-levels of frustration. Our code is available online at this https URL.

Secure and Safe IIoT Systems via Machine and Deep Learning Approaches

Abstract: This chapter reviews security and engineering system safety challenges for Internet of Things (IoT) applications in industrial environments. On the one hand, security concerns arise from the expanding attack surface of long-running technical systems due to the increasing connectivity on all levels of the industrial automation pyramid. On the other hand, safety concerns magnify the consequences of traditional security attacks. Based on the thorough analysis of potential security and safety issues of IoT systems, the chapter surveys machine learning and deep learning (ML/DL) methods that can be applied to counter the security and safety threats that emerge in this context. In particular, the chapter explores how ML/DL methods can be leveraged in the engineering phase for designing more secure and safe IoT-enabled long-running technical systems. However, the peculiarities of IoT environments (e.g., resource-constrained devices with limited memory, energy, and computational capabilities) still represent a barrier to the adoption of these methods. Thus, this chapter also discusses the limitations of ML/DL methods for IoT security and how they might be overcome in future work by pursuing the suggested research directions.

Integrated Cyber Safety & Security Management System: Industry 4.0 Issue

Abstract: The aim of the article is to enhance the conventional concepts of enterprise safety and security management by combining this two interrelated features in one philosophical and managerial conception. Overall protection of the enterprise, where critical data assets are stored, processed and transmitted is impossible without uniting security and safety systems properties. It is especially essential nowadays with intensively evolving Industry 4.0 in developing countries like Ukraine and rapidly developing Industry 5.0 in developed countries with digitalized societies. A three-level hierarchical integrated model of the architecture of a functional safety and security management system that meets the requirements of international standards is shown. It includes principles of the control law and cybernetic control.