ArticlesACP Model for Vehicle Monitoring Based on CPS
Cyber-physical systems (CPS) integrate cyberspace with the physical world through a network of interrelated elements, such as sensors and actuators, and computational engines. A CPS must also have the ability to overcome the uncertainties inherent to the physical system and its environment. However, an inspection of the real-time surrounding surveillance is highly dynamic in a regularly changing condition. The dynamic situation inflicts significant losses and potentially destructive impacts on people, equipment, and the environment due to inevitable abrupt malfunctions. For a CPS to deliver both availability and safety, it is essential to have comprehensive system prognostics and condition care in order to observe and verify the physical system’s life cycle. This study proposes an approach to monitoring the integrated vehicle condition in a CPS based on the ACP (awareness, cognitive, and prediction) model. The proposed method is as follows: appropriate digital models are structured through constant monitoring using a combination of physical multi-sensors; a cognitive interaction is formulated between the physical world and cyberspace with data orchestration; and abnormal signs are diagnosed and faults (fault, error, failure, accident) are predicted by acquiring an understanding of the current situation. This work is designed to ensure the safe operation of complex systems, and is considered useful in developing a realistic variety of cyber-based monitoring systems.
Cyber-Physical System, Awareness, Cognition, Diagnostic and Prognostic, Prediction, Health Management
Based on modern technologies, smart systems comprise numerous physical elements (sensors/actuators) and construct a connected space to services for transmitting information [1–3]. A connected space leads to interaction between the physical world and the cyber world, thus facilitating monitoring and independent control of a real physical entity. Cyber-physical systems (CPS) include the integration of computing, communication, monitoring, and management processes to provide an efficient way of exploiting interdependent behaviors when working with physical methods in diverse applications [4, 5]. Physical systems have various risk factors, such as highly dynamic environmental factors and system failures. These factors are a significant cause of system malfunction that can inflict severe damage or catastrophic consequences on people, equipment, and the environment. In particular, malfunctions of safety-critical systems used in the aero-spatial, automotive, and railroad sectors are enormously damaging because they are directly related to social confusion and safety accidents . Due to the great complexity of safety-critical systems, there are significant issues concerning diagnosis of the current condition and the reliability of fault detection which need to be resolved.
Researchers on safety-critical systems highlight the challenges encountered in developing condition monitoring systems and how to solve these challenges [7–10]. The study  provides an overview of the prognostics and health management (PHM) practices used in the production system, including problems, requirements, and methods. Another study  outlined a multi-layered diagram representation of a CPS in order to identify critical scenarios and determine the sources of risk, by considering humans’ role in wrong decisions and their potential influence as sources of risks. One study  described a completely unsupervised and partially interactive CPS-based PHM methodology that relies on streaming data collected from sensors and performs computations either at the edge or in the cloud. The study  discussed the shortcomings found in the field of integration vehicle health management (IVHM), especially in condition monitoring at the vehicle level, and how to apply argumentation to address some of those shortcomings. As a result, the implementation of safety-critical systems based on CPS should focus on understanding complex physical systems, deep integration, and real-time interactions between the physical and virtual worlds.
This paper proposes CPS-based integration care monitoring of the ACP (awareness, cognition, and prediction) model applied to vehicle systems. The ACP model uses the A (awareness)–C (cognition)–P (prediction) process to solve the requirements mentioned above. Fig. 1 shows integration monitoring for system prognostics and condition care based on the ACP model. The awareness process is an operation for structuring appropriate digital models by accurately recognizing physical objects and their surrounding environments. The cognition process, through interaction between both systems, detects the situation of the physical model. The prediction process issues an action for the diagnostics’ faulty condition based on the detection of situations. Based on the vehicle systems, each process provides a detailed description of its behavior and needs.
Fig. 1. Monitoring for system prognostics and condition care based on the ACP model.
This paper’s main contributions are prognostics and condition diagnosis in order to observe and verify the physical system’s life cycle using cyber-based monitoring systems. The rest of the paper proceeds as follows: Section 2 describes state-of-the-art related works; Section 3 presents an overview of the characteristics of the ACP model including the full architecture; Section 4 details the ACP properties approach for vehicle systems; Section 5 presents the discussion; and, finally, Section 6 presents the conclusion.
The availability of CPSs has a significant impact on the growth of large distributed systems in such areas of research as smart factories, intelligent/autonomous vehicles and communications infrastructures. A CPS can provide safe, reliable, and dynamic real-time interaction with physical systems [11, 12].
As the complexity and automation of physical systems continue to increase, there is a need for a safety-critical system based on CPS that provides availability and safety. This is because the degradation of performance can lead to significant losses and even catastrophic consequences for people. PHM is used to proactively monitor system health, perform diagnostics and prognostics, and provide operational maintenance strategies. The PHM approach typically entails performing multiple tests that simulate fault conditions in order to collect a training set for model training. There are two main problems with this approach. First, data collection is a time-consuming task and requires prior knowledge of all fault sources. Second, the existing learning model is trained offline and cannot adapt to the equipment’s real-time state . There have been some studies of large amounts of data using several machine learning (ML) and artificial intelligence (AI) algorithms for function extraction, diagnosis, and prognosis [, ]. Since the CPS-based system must detect real-time situations, the PHM-based approach is unsuitable as it does not consider accurate interaction between systems.
The IVHM technology uses a vehicle’s sensor data to perform diagnostics and prognosis in order to evaluate and predict system conditions . The IVHM technology enables condition-based maintenance (CBM) of vehicles like aircraft. CBM is a program that provides maintenance plans according to the condition of a given vehicle. The method are pre-modeling a set of condition-rules related to scenario situations and modulate rules within a framework. It measures product performance in real time and replaces traditional time-based programs such as regular inspections and preventative maintenance. However, the application of reliability-based designs is limited because it cannot explain all operational situations.
A CPS contains various sensitive devices and an actuator technology that can be connected to create a network. Enhanced by network protocols, the system can perform the acquisition of real physical information with the processing required to support knowledge of how the system works internally.
Each of the above approaches has many options for performing diagnostic and predictive maintenance. Table 1 shows a comparison of the relevant works. It shows that performance is dependent on training data and may deteriorate when the system is operating under unknown conditions or in novel failure modes. In the present study, the ACP can integrate the system’s physical understanding to describe the physical system’s dynamic behavior. Through orchestration and synchronization between virtual models and physical objects are supported interaction with the environment in detecting and diagnosing problems.
Table 1. Comparison of condition monitoring in a CPS
||Interaction between physical and cyber model
|, , 
|, , 
The safe systems based on a CPS should focus on a complete model that awareness the physical structure, provides the interaction between the physical and cyber systems, and predicts the diagnosis fault conditions. This paper proposes a comprehensive infrastructure of integration vehicle prognostics and a condition monitoring and predictive maintenance system-based ACP model. The ACP detects complex physical systems, provides knowledge of the surrounding environment, and monitors condition diagnosis and fault detection based on perceived situations.
ACP Model for CPS
A CPS is essential to achieve deep integration and real-time interaction between the physical and virtual worlds in order to improve integrated condition monitoring for complex systems. To this end, the recognition, understanding, connection, and prediction diagnostic (operational functions/capacities) of CPS-based monitoring systems are becoming increasingly important. This chapter presents an overview of the ACP model. The ACP provides a way of monitoring systems that recognizes the environment, understands both the environment and the situation through deep integration between physical and cyber objects, and then interacts with them in real time. Fig. 2 shows the ACP architecture for integrated monitoring systems in the CPS.
The integration monitoring systems in the CPS, as shown in Fig. 2, are divided into the physical world and the CPS. The physical level is a multi-sensor data acquisition and pre-processing module that provides the source from the physical aspect expressed by sensors and activators. Each sensor and actuator includes computing faculties and capabilities, which can be considered as “things”, and the sensors at the physical level may be part of a wide range of CPS systems. For instance, they may be part of a data hybrid collection system that includes low-power radio communications and various feedback actuators. The CPS includes real-time tracking and real-time monitoring methods to determine whether arrangements should be controlled and whether to change a current action if it is unsafe. The CPS goal is to prognostics on whether the situation is dangerous or not. To observe and verify the physical system, the ACP includes three processes: awareness, cognition, and prediction of the structure of an aware-safety system. The ACP provides an interaction between physical and cyber components, contextual identification, prediction condition diagnosis, and decision-making. The following paragraphs describe each process of the ACP.
Fig. 2. The architecture of the ACP model for integrated monitoring systems in the CPS.
Perception and understanding of the physical system are probably among the most necessary components of the CPS-based safety system. The difficulty of the perception environment is that physical objects are generally parts of the entire system and move over time within the surrounding environment. Both the perception and the modeling surrounding such mobility situations are complex and must develop carefully, as only then can a suitable digital model of the physical system be created. This measurement is different from the simulation. The CPS’s digital model should reflect and control each object’s components and the surrounding environment of a complex system. The awareness process of the ACP constructs an appropriate digital model that accurately recognizes physical objects and surroundings. The awareness method includes hybrid sensing, fusion, and localization and implements self-awareness information (see Fig. 3).
Hybrid sensing simultaneously collects data from the physical object and the surrounding environment through various sensors and actuators. Hybrid sensing is a series of three stages: data acquisition, preprocessing, and data mining. Based on the hybrid sense, the quality and rate of data are improved, allowing data fusion. Here, “sensor fusion” refers to the process of combining data from several different physical things to produce more accurate proportions. Localization determines the physical object’s position and the dimensions of its movements. This self-awareness using to construct appropriate digital models by accurately recognizing physical objects and their surrounding environments.
Fig. 3. Awareness process.
A CPS has various unpredictable variables because the virtual world and the physical world work according to different mechanisms. For example, in terms of different time systems and units of delay, the physical world deals with delays due to continuous time (seconds, milliseconds) and the movements of physical devices. In contrast, the virtual world deals with discrete-time, network, and processing delays. The difference in the concept of time becomes a potential risk factor that could cause an accident. To ensure a CPS’s safety and reliability by reflecting these characteristics necessitates interactions between the physical and virtual worlds. The cognition process supports the deep integration of real-time interactions between the physical and virtual worlds to detect situations. There are two essential functional modules in this layer: orchestration and synchronization. Fig. 4 shows the cognition process.
Orchestration is a task involving the categorization and objectification of each physical functionality of the system with virtual/cyber objects. Synchronization performs a one-to-one mapping relationship based on collaboration and knowledge-sharing between physical and virtual objects.
The prediction process issues decision actions based on the detection of abnormal signs or a situation involving faulty conditions. In most accidents, hidden faults appear at unexpected moments during system operation. Such manifestations occur as a continuous propagation process explained by the chain-
Fig. 4. Cognition process.
link of “fault-error-failure-accident”. The prediction process logically identifies and analyzes the causes of accidents that may occur during operation based on this chain-link of the system. Fig. 5 shows the prediction process of the ACP.
The fault detection engine deduces risk factors using forward and backward analysis. Forward analysis is a method of predicting and preparing for possible accidents in a system based on the well-known “safety case-based” scheme. A safety case is an evidence formation of distinct arguments based on prior experience. Thus, whereas the forward analysis method predicts an accident according to the chain-link of “fault-error-failure-accident”, backward analysis is a way of deducing possible accidents in the current situation. In other words, it detects system changes caused by a dynamic environment and predicts possible incidents in the current situation. Also, backward analysis derives risk factors for potential accidents. The method of backward analysis indicates a risk factor according to the backward chain-link of “accident-failure-error-fault”.
Fig. 5. Prediction process.
A-C-P Process for Vehicle Monitoring
The CPS-based system for vehicles is made more complex by the use of various environmental sensors, such as radar, cameras, and ultrasonic sensors covering all possible angles, actuators, and control. In addition, to measure the working conditions and movements of a vehicle, steering angle sensors, speed sensors, gyroscopes, and acceleration sensors can be installed. If any detectors or actuators are temporarily defective during operation due to environmental ambiguity, they may read data incorrectly. This information is essential because if a malfunction occurs in these physical objects during their autonomous movement, the vehicle may deviate from its route or be unable to perform the correct operation necessary for control, which could lead to an accident. To prevent such accidents, the analysis of data from the physical systems of different detectors and actuators should provide a solution for understanding and prediction in the real world.
This section provides an overview of the vehicle monitoring in a CPS through the ACP model. The ACP model uses the A (awareness)–C (cognition)–P (prediction) process for integration monitoring to prognostics and diagnosis of physical systems. Fig. 6 represents the process of A–C–P processing.
The CPS-based ACP framework includes two symmetrical worlds: the cyber world and the physical world, as shown in Fig. 6. The physical world is mainly composed of two main modules are sensors and actuators. Sensors can detect the physical components and dynamic state of physical features (such as changes in operating processes or vehicle speed) and acquire data from their surroundings (such as traffic conditions and road infrastructure). The actuator is a machine component responsible for activating and controlling mechanism parameters through a control signal operation. The physical side of the CPS uses sensors and actuators to monitor and control the physical world.
In cyberspace, the CPS consists of an ACP model that recognizes the environment, detects situations through deep integration between physical and cyber objects, and issues an action for a detected problem. The ACP includes three main processes: awareness, cognition, and prediction, as shown in Fig. 6. The awareness process fuses and localizes physical data collected through hybrid sensing to build an appropriate virtual model of the physical world. The cognitive process supports the integration of interactions between the physical and virtual worlds to detect situations. The prediction process issues decision actions based on the detection situation. The details of each component are given in the following subsections.
Fig. 6. A–C–P process to prognostics and the condition diagnosis.
The awareness process is the first and most important process by which an appropriate digital model is built in order to reproduce each physical object in a virtual object, as shown in Fig. 7.
The hybrid sensing module acquires physical components, physical dynamics (physical object behaviors), and environmental dynamics (physical object’s surrounding environment) through the actuators and sensors. These sensors’ data can extract various physical parameters, detect events or changes in the environment to help understand vehicle status, and surround surveillance. Data of physical objects and behavior can extract from vehicle inside sensors and OBD. Environment data provide sensor devices such as radar, lidar, and camera. The hybrid identifies all features of the physical component since different characteristics are needed depending on each object. The addressing and naming classes include such features as PhysicalComponentID, PhysicalComponentType, PhysicalComponentName, PhysicalEnvironment, Timestamp, and other descriptions. Physical object includes such features as ObjectID, ObjectType, ObjectName, ConnectionAdapter, sensors, and actuators. The sensors and actuators include the features of ID, name, type, and value. They need to be recorded and integrated on a system platform for continuous connectivity to transfer collected data.
Fig. 7. ACP construction a digital model based on the awareness process.
Fusion combines external sensors’ data in the surrounding environment, such as traffic or road information collected via RSUs (roadside units) or a TS (transport system) that may distribute available software updates for vehicles in a specific area. Localization determines the physical object’s position, the dimensions of its movement route, and its destination. In combination with GPS and INS data, vehicle location can consider using Bayesian filtering procedures.
Self-awareness creates an appropriate digital model to reproduce each physical object as a virtual object based on the above hybrid data. Virtual objects represent concrete physical objects. The awareness process of the ACP can be defined as follows:
In the above formula (1), VOI is a set of virtual objects information, including all virtual objects (VO) representing physical objects (P0) in unique VOid
. The VOinfo
includes detailed descriptions such as shape, properties, and processing capacity. VOI and VO are formalized as formulas (2) and (3):
The VOB in formula (1) is a set of physical object behaviors (POB) reproduced in a virtual form that can perceive its own running status and environment. The resource scheme is formalized as follows:
is a unique ID of the VO behavior; VOBtype
is a classifier of the behavior type; VObset
is a set of working states including angle A, energy E, speed S, and vibration V; VObstate
represents the current state, including the action (true) and non-action (false); and VObinfo
is the detailed information of a resource, such as a responsible operator or model.
The VOE in formula (1) is a virtual object environment describing the static and dynamic attributes of the physical object’s surrounding environment (POE) in the physical world. Connected vehicles include a complex set of built-in smart sensors, such as radar, LiDAR, cameras, ultrasound, and related sensors for V2X communication, and provide the real-time data required to build knowledge of its surroundings at any point in time.
Through the object virtualization process, the ACP captures the physical modules’ attributes and operations (sensors and actuators) in virtual objects, virtual resources, and surrounding environments, and then stores them in a self-awareness database. As the vehicle status, behavior, and condition of the physical world are dynamically changing, the awareness process is continuously updated, creates, uses, and stores all data until the vehicle reaches its destination.
The cognition process then uses self-awareness information and detects the situation of cognition objects (CO_situational). Cognition includes orchestration and logical synchronization, which enable integration between the physical and virtual models in order to understand and detect situations. The situation of cognition objects can be formulated as follows:
Orchestration provides a cognitive-link (⨂) of the VO with the dynamic behavior data of each physical object that is represented in the VOB. The orchestration is formalized as:
Synchronization provides a mapping relationship between VO and PO. The mapping relationship between the physical and digital worlds of the CPS provides a one-to-one correspondence.
Synchronization is formalized with the operator (
), which denotes a one-to-one mapping relationship between each object in the cyber and physical worlds.
Through the cognitive process, the ACP can obtain the current state of physical objects and detect the situation. In this paper, a detected situation means the detection of abnormal signs and failure conditions.
The prediction process issues an action for a detected situation. To make a decision, the prediction process uses a decision tree model based on a critical-safe case. The critical-safe case represents the three-step actions of risk factors received through the learning model in the predictive engine. The prediction engine logically identifies and analyzes the causes of accidents. The accident reasons arising during operation are determined based on the chain-link “failure-error-failure-accident”. The first step involves a “warning” decision action, where chain-link is fault-error; risk factors are not dangerous, but need to be reviewed. The second step involves a “recommendation” decision, where chain-link is fault-error-failure; these risk factors can lead to a hazardous situation. The third step is decision “decree”, where chain-link is fault-error-failure-accident; the risk factors are critically dangerous.
The predictive engine deduces risk factors using forward and backward analysis. In a dynamic environment, time-series data can contain various patterns. Since all patterns cannot be predicted and defined in advance, the forward analysis includes a three-stage rule algorithm to detect essential system patterns. The first algorithm uses outlier detection to capture the moment of anomalous data. Based on CPD (Pattern Change Detection), the second algorithm detects the changing pattern. The third algorithm uses an anomaly detection technique to detect the moment when the detection frequency appears.
Based on the change in the state of the environment, the backward analysis of a variety of probability accidents calculated using the Probability Theory for Time Series model uses to infer an accident’s probability. The prediction level structures a vehicle-aware safety system by detecting the static and temporal signals of behavior and then learning probabilistic conclusions.
Based on the interaction and integration between the physical and cyber worlds, CPS-based monitoring systems can diagnose faults in complex physical systems. This paper focuses on integrating vehicle condition monitoring into a CPS based on the ACP model. The ACP is a model that provides recognition, understanding, connection, and prediction diagnostic based on a vehicle’s condition and its surrounding environment. Fig. 8 shows a scenario of how accidents can be prevented with an ACP in the event that a sinkhole appears suddenly.
Fig. 8. ACP Model for vehicle monitoring based on the CPS.
The awareness process (marked as number ① in Fig. 8) in the ACP provides the structuring of appropriate digital models by accurately recognizing physical objects and their surrounding environments. The digital model reproduces each physical object as a virtual object. Virtual things are specific physical objects. The cognitive process (marked as number ② in Fig. 8) supports the integration of interactions between the physical and virtual worlds in order to detect situations. Through deep integration, the detection situations occurring in the physical environment will be derived in real time. In the scenario, when a sinkhole appears suddenly in a dynamic environment, cognition provides an event based on the diagnosis of abnormal signs of a faulty condition through detection situations. The prediction process of the ACP (marked as number ③ in Fig. 8) issues a decision/prediction action based on the risk situation. The prediction process derives the three-step actions of risk factors by identifying the current situation’s danger level based on a critical-safe case. In the scenario, through the vehicle’s current position, relative position, distance, and speed the predicted engine execution step is judged as “critically dangerous” and action is taken to prevent an accident.
As indicated in the scenario shown in Fig. 8, the physical system has many risk factors, such as dynamic environmental factors or system failures. These factors can cause severe damages and/or fatal consequences for people and the environment. In particular, malfunctions of the safety-critical systems used in a vehicle are directly linked to social confusion and safety accidents, resulting in enormous damages. In this regard, the application of the ACP model to traffic management monitoring can prevent dangerous accidents.
The CPS is a new direction which allows physical systems to efficiently use interdependent behavior when working with physical methods of implementing safety-critical strategies. Safe systems based on a CPS should be focused on a complete model that recognizes a physical structure, detects a situation through the interaction between the physical world and the cyber system, and decisive action based on a fault conditions factor of diagnosis.
This paper presents integration monitoring based on the ACP model. ACP is a design methodology that comprises a way for systems to provide a recognition environment, to understand the environment and the situation, and to connect physical objects with cyber objects. Also, ACP can integrate a physical system understanding so as to describe a physical system’s dynamic behavior through orchestration and synchronization between virtual models and physical objects. Research is currently underway to collect real-time data for testing using RC cars and simulations “safety case-based” situations. The cognitive layer conducts orchestration and cognition connection studies to connect physical entities with cyber systems based on real data in simulated situations. Future research will consist of a real-time awareness environment based on accurate data and a description of the cognition section connecting it to the cyber network.
Conceptualization, KS, YI. Funding acquisition, KS. Investigation and methodology, KS. Project administration, KS, YI. Resources, KS. Supervision, YI. Writing of the original draft, KS. Writing of the review and editing, KS, YI.
This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (No. 2018R1D1A1B07047112).
This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (No. NRF-2019R1I1A1A01064054).
The authors declare that they have no competing interests.
Name: Svetlana Kim
Research Professor of IT Engineering, Sookmyung Women's University
She received her BS and MS degrees in multimedia science from the Sookmyung Woman’s University in 2005 and 2007. She received her PhD in Distribution system at the Sookmyung Woman’s University in 2017. Since 2008, she has been a visiting professor at her alma mater. Her research interests are in the area of ubiquitous computing, distributed middleware, mobile computing, MPEG-21, cloud computing, ELearning, N-screen standardization, synchronization, digital communication systems, BigData, Fusion sensors, AI, Edge computing, CPS (Cyber-physical system), situation awareness and Hybrid IoT.
Name: YongIk Yoon
Professor of IT Engineering, Sookmyung Women's University
He received his BS in Statistics from the Dongguk University in 1983 and MS degree in computer science from Korea Advanced Institute of Science and Technology (KAIST) in 1985. From 1985 to 1997, he served as senior researcher at Electronics and Telecommunications Research Institute (ETRI) in following research projects; the research project of development environment of exchange system, the TDX-10 exchanger development project, the mobile communication development project, and ATM exchange development project. He received his PhD in multimedia science and distribution system from KAIST in 1994. Since 1998, he has been a professor of Sookmyung Woman’s University. His interests include middleware, smart services, IoT, situation awareness, embedded system, ubiquitous computing, distributed system, real-time processing system, real-time OS/DBMS and BigData.
 E. M. Rubio, R. P. Dionisio, and P. M. B. Torres, “Industrial IoT devices and cyber-physical production systems: review and use case,” in Innovation, Engineering and Entrepreneurship. Cham, Switzerland: Springer, 2018, pp. 292-298.
 D. Lee and J. H. Park, “Future trends of AI-based smart systems and services: challenges, opportunities, and solutions,” Journal of Information Processing Systems, vol. 15, no. 4, pp. 717-723, 2019.
 J. H. Park, M. M. Salim, J. H. Jo, J. C. S. Sicato, S. Rathore, and J. H. Park, “CIoT-Net: a scalable cognitive IoT based smart city network architecture,” Human-centric Computing and Information Sciences, vol. 9, article no. 29, 2019. https://doi.org/10.1186/s13673-019-0190-9
 C. Zhang, X. Xu, and H. Chen, “Theoretical foundations and applications of cyber-physical systems: a literature review,” Library Hi Tech, vol. 38, no. 1, pp. 95-104, 2019.
 N. Y. Kim, S. Rathore, J. H. Ryu, J. H. Park, and J. H. Park, “A survey on cyber physical system security for IoT: issues, challenges, threats, solutions,” Journal of Information Processing Systems, vol. 14, no. 6, pp. 1361-1384, 2018.
 Telecommunications Technology Association (TTA), Guideline for assuring the trustworthiness of Cyber-Physical Systems (CPS) Part 1: CPS Accident Analysis Model, TTAK.KO-11.0268-Part1, 2019.
 G. W. Vogl, B. A. Weiss, and M. Helu, “A review of diagnostic and prognostic capabilities and best practices for manufacturing,” Journal of Intelligent Manufacturing, vol. 30, pp. 79-95, 2019.
 N. H. Carreras Guzman, M. Wied, I. Kozine, and M. A. Lundteigen, “Conceptualizing the key features of cyber‐physical systems in a multi‐layered representation for safety and security analysis,” Systems Engineering, vol. 23, no. 2, pp. 189-210, 2020.
 F. Calabrese, A. Regattieri, L. Botti, and F. G. Galizia, “Prognostic health management of production systems: new proposed approach and experimental evidences,” Procedia Manufacturing, vol. 39, pp. 260-269, 2019.
 C. M. Ezhilarasu, Z. Skaf, and I. K. Jennions, “The application of reasoning to aerospace Integrated Vehicle Health Management (IVHM): challenges and opportunities,” Progress in Aerospace Sciences, vol. 105, pp. 60-73, 2019.
 D. Cogliati, M. Falchetto, D. Pau, M. Roveri, and G. Viscardi, “Intelligent cyber-physical systems for industry 4.0,” in Proceedings of 2018 1st International Conference on Artificial Intelligence for Industries (AI4I), Laguna Hills, CA, 2018, pp. 19-22.
 Y. Liu, Y. Peng, B. Wang, S. Yao, and Z. Liu, “Review on cyber-physical systems,” IEEE/CAA Journal of Automatica Sinica, vol. 4, no. 1, pp. 27-40, 2017.
 S. Martin-del-Campo and F. Sandin, “Online feature learning for condition monitoring of rotating machinery,” Engineering Applications of Artificial Intelligence, vol. 64, pp. 187-196, 2017.
 R. Liu, B. Yang, E. Zio, and X. Chen, “Artificial intelligence for fault diagnosis of rotating machinery: a review,” Mechanical Systems and Signal Processing, vol. 108, pp. 33-47, 2018.
 F. Castano, S. Strzelczak, A. Villalonga, R. E. Haber, and J. Kossakowska, “Sensor reliability in cyber-physical systems using internet-of-things data: a review and case study,” Remote Sensing, vol. 11, no. 19, article no. 2252, 2019. https://doi.org/10.3390/rs11192252
 N. Kumar, M. Singh, S. Zeadally, J. J. Rodrigues, and S. Rho, “Cloud-assisted context-aware vehicular cyber-physical system for PHEVs in smart grid,” IEEE Systems Journal, vol. 11, no. 1, pp. 140-151, 2017.
 Z. Wang, Y. Wu, and Q. Niu, “Multi-sensor fusion in automated driving: a survey,” IEEE Access, vol. 8, pp. 2847-2868, 2019.
 M. Canizo, A. Conde, S. Charramendieta, R. Minon, R. G. Cid-Fuentes, and F. Onieva, “Implementation of a large-scale platform for cyber-physical system real-time monitoring,” IEEE Access, vol. 7, pp. 52455-52466, 2019.
About this article
Cite this article
Svetlana Kim and YongIk Yoon*, ACP Model for Vehicle Monitoring Based on CPS, Article number: 11:05 (2021) Cite this article 17 Accesses
- Recived25 March 2020
- Accepted27 December 2020
- Published29 January 2021
Share this article
Anyone you share the following link with will be able to read this content:
Provided by the Springer Nature SharedIt content-sharing initiative
- Cyber-Physical System
- Diagnostic and Prognostic
- Health Management