Hazard-intense applications of cyber-physical systems (CPSs), such as the evacuation of a building on fire, require personalized informing on the basis of a real-time assessment of dynamic context. In this paper, a context-dependent message construction mechanism (CD-MCM) is proposed with the objectives (i) to inform people about the emergence and development of a situation unsafe for them, and (ii) to instruct them what they have to do according to an adaptively computed action plan. To achieve this, the concepts of 'situation' and 'impact indicator' have been introduced in order to facilitate the computation of personalized action plans and to send messages about the level of danger and the requested actions. In both activities, the inferred implications of situations are used as the basis of informing the involved people. The messages are adapted to actual situations. In addition, the concept of 'relevance indicator' was utilized to assess the significance of the standing situations for the concerned people in a quasi-real-time manner. The level of danger was evaluated for each person by totaling the values of the situation-related relevance indicators. This was also used to select the proper message templates from the predefined alternatives in the process of message construction. The personalized messages were generated based on the chosen message template and the various message components describing concerned situations or providing instructions. The CD-MCM was validated in a simulated indoor fire evacuation guiding application. In the practical evaluation of the quality of the generated messages, a sample of test people was involved. The results of the evaluation show that the messages generated by the proposed CD-MCM lead to more effective messaging about the personal context and the expected actions than the messages constructed by using static context information only. The reason is that the proposed template-based message construction mechanism facilitates the appropriateness as well as the articulation of the contents of context-sensitive personalized messages.

@en

An approach to repairing surface defects of metal parts is proposed, which includes a combined application of (i) groove machining, (ii) wire and arc additive manufacturing (WAAM), and (iii) finishing machining. The completed analysis revealed that (i) the inclination angle of the groove to be machined is strongly influenced by the manufacturing parameters of the WAAM process, and (ii) the WAAM process models designed for fabricating parts on a flat substrate are not appropriate for filling grooves. To overcome these issues, this research investigated the range of variation of the proper inclination angle of the groove. A mathematical model was developed to determine the manufacturing parameters of WAAM that result in a proper filling of the groove. The effectiveness of the proposed fundamentals was investigated in a case study. The experimental results showed that using the proposed approach and the chosen manufacturing parameters resulted in a complete filling of the machined groove. The fabrication error of the main part of the repaired region before the finishing machining was less than 0.3 mm, while the ‘buy-to-fly’ ratio of the deposited material was 92.1%. The proposed approach for morphological repair lends itself to a computer-aided automatic part repair process.

@en

Context is interpreted as a body of information dynamically created by a pattern of entities and relationships over a history of situations. Computational handling of dynamically changing contexts and the consideration of rapidly changing situations in awareness building, situated reasoning, and proactive adaptation of smart cyberphysical systems has been recognized as an important research phenomenon. The main reason is that there are many real-life processes whose smart control and self-* behavior require quasi-real time processing of context information. Though processing time-varied context information has been addressed in the literature, domain-independent solutions for reasoning about time-varying complex and critical activity scenarios are scarce. Thus, explicit generation and utilization of dynamic context semantics in smart behavior of informing cyber-physical systems (I-CPSs) is a frontier endeavor. @en

Computational mechanisms (complex structures of functionally connected algorithms) are developed very frequently by both the academia and the industry for specific applications, but the transferability to other applications is seldom addressed and systematically tested. This raises the need for application validity of constructive computational methodologies (CCMs). First, this background research paper clarifies the fundamental concepts related to application validation. Applicability validation focusses on the indicators of appropriateness with regards to a particular purpose. Then, it surveys the various approaches of application validation based on the publications available in contemporary literature. Its main finding is that applicability validation of CCMs seems to be a stepchild of academic research. The same applies to the industrial exploration of the applicability of CCMs tailored to a narrow family of applications to a wider range of applications.

@en

Mathematical modelling of the shape-forming process of multi-layer multi-bead parts dealing with (i) a single layer, (ii) inclination angles and (iii) various slopes, is carried out. In the layers-overlapping model, material shortage areas are generated at the edges of layers. To solve the problem, the deposition amount of beads at the edges of the second and above layers should be modified, in association with the parameters applied to fabricate the slope. Various basic inclined components were deposited to validate the proposed mathematical formulation. Findings from the basic experiments were: (i) depositing additional material to compensate for the material shortage areas is necessary for realizing the designed geometries of the layers, (ii) depositing the bead at the layer edges alongside an already-deposited neighboring bead enables a better shape-forming in the case of negative slopes and (iii) the deposition order of beads in a layer has little influence on the shape formation of positive slopes. A complex part was also fabricated as a case study to validate the model and findings in the real-life context. Practical issues with regard to using the model and the solutions to the issues were discussed.

@en

Robotic wire and arc additive manufacturing (WAAM) systems are required to provide predictable and efficient operations to fabricate solid metallic parts with high morphological fidelity and geometric accuracy. Since the metallic parts are fabricated based on a layer-by-layer principle, the interactions between the neighboring beads and layers strongly influence the geometric accuracy of the fabricated part. The layers-overlapping process has been studied and a traditional layers-overlapping model (T-LOM) has been published in the literature. This paper proposes a layers-overlapping strategy (LOS), based on which a revised layers-overlapping model (R-LOM) was proposed for the fabrication of multi-layer multi-bead (MLMB) components with homogeneous layers. A mathematical model for layers-overlapping is presented, which considers the material shortage areas at the edges of the layers. This is important since the material shortage areas result in a situation that the component width is smaller than the expected value. In addition, they will be accumulated when multiple layers are overlapped through normal unidirectional parallel (NUP) paths. The proposed LOS addresses two aspects: (i) the deposition amount of the first bead and the last bead in the lap layers should be increased and (ii) the deposition position of the first bead and the last bead in the lap layers should be moved towards the edges with a given offset distance. Validation experiments were designed and conducted to test the proposed concepts and models. The experimental results indicated that (i) the R-LOM enables the MLMB components to achieve the expected width and (ii) for components deposited with NUP paths, the R-LOM eliminates the effect of accumulation of material shortage areas on the first bead and increases the surface flatness.

@en

Wire and arc additive manufacturing (WAAM) is a competitive technology for fabricating metallic parts with complex structure and geometry. It enables the fabrication of multi-layer multi-bead (MLMB) parts. The basis of planning the deposition paths is the beads overlapping model (BOM). The existing overlapping models consider only the geometric area of adjacent beads, but ignore the spreading of the melted weld beads. The objective of the research was to develop an enhanced BOM (E.BOM) for WAAM, which takes the spreading of the weld beads into consideration. A deposited bead spreads to the already deposited neighboring bead and as a consequence, its center point deviates from the center point of the fed (to be melted) wire. Experiments were designed to explore the relationships between the geometries of the beads, and the offset distance between the center of a weld bead and the center of the fed wire. An artificial neural network was used to predict the offset distance of a certain weld bead based on the results of the experiments. In addition, a reasoning algorithm was implemented to calculate the optimal distance between the centers of adjacent deposition paths in order to achieve a planned center distance between adjacent beads. This enables the control of the actual center distance of the adjacent beads according to an expected value. The E.BOM has been tested by validation experiments. On the one hand, it improves the surface flatness of layers of MLMB parts produced by WAAM. On the other hand, it prevents formation of defects inside the parts.

@en

Authors belonging to different institutions (‘schools’) of cyber-physical systems (CPSs) research and development report on largely different objectives, underpin their work with different theories and methodologies, and target characteristics which can actually better characterize other categories and families of engineered systems. This has resulted in an ontological chaos. Therefore, our research addressed the question: What exists in the form of past, current and future CPSs? Our hypothesis has been that we can have an ordered picture on the landscape of CPSs by introducing the notion of system generation. Generation is a structural term defined as a ‘technological/engineering cohort’ of different individual manifestation of systems that reflect genotypic features of ancestor systems belonging to the same category, but deviates from them with regards to their phenotypic features. Based on our literature findings, we have defined five generations of CPSs, which could be differentiated based on: (i) the level of self-intelligence, and (ii) the level of self-organization. The zeroth generation includes look-alikes and partial implementations of CPS. The 1G-CPSs include systems with self-regulation and self-tuning capabilities, while the 2G-CPSs are capable to operationalize self-awareness and self-adaptation. The 3G-CPSs are equipped with the capabilities of self-cognizance and self-evolution. According to our reasoning model, only the fourth generation of CPSs is supposed to achieve self-consciousness and self-reproduction in the form of system of systems. The paper analyses the major paradigmatic characteristics of these generations. It also provides an outlook to the trends that may have strong influence on the introduced generations of CPSs.@en

There are many real life processes whose smart control requires processing context information. Though the issue of processing varying context information has been addressed in the literature, domain independent solutions that can support reasoning and decision making according to time-varying process scenarios in multiple application fields are scarce. This paper proposes a method for dynamic context computation concerning spatial and attributive information. Context is interpreted as a body of information dynamically created by a pattern of entities and relationships over a history of situations. Time is conceived as a causative force capable of changing situations, and acting on people and objects. The invariant and variant spatial information is captured by a two-dimensional spatial feature representation matrix. The time-dependent changes in the context information are computed based on a dynamic context information management hyper-matrix. This humble but powerful representation lends itself to a quasi-real time computing and is able to provide information about foreseeable happenings over multiple situations. The paper uses the practical case of evacuation of a building in fire both as an explorative case for conceptualization of the functionality of the computational mechanism and as a demonstrative and testing application. Our intention is to use the dynamic context computation mechanism as a kernel component of a reasoning platform for informing cyber physical systems.@en

Informing cyber-physical systems (I-CPSs) are designed to accomplish sensing, reasoning and informing activities in dynamic context. In order to simplify and accelerate the design and implementation process of multiple context-aware ICPSs, we are developing an information sensing, computing and actuating (SCA) platform that can be used as a central module of these systems. This paper presents the concept of a SCA platform. The functionality of the platform includes development of context-dependent strategies to adapt the sensing, reasoning and informing behaviors of the platform to various dynamic contexts. There are four constituents of the platform: (1) a generic kernel, (2) built-in elements, (3) add-on components, and (4) system interfaces. The paper also discusses both the internal and external integration mechanism of the SCA platform, which can be customized according to the needs of specific I-CPS applications by extending the generic kernel with various functional built-in elements and add-on components. The feasibility and applicability of the platform have been tested through a case study: an indoor fire evacuation guiding system. The proposed platform provides a useful package of functionalities, alleviates the burden of developers, and speeds up the development of applications specific context-aware I-CPS.@en

There are many real-life processes whose smart control requires processing context information. Though processing time-varied context information is addressed in the literature, domain-independent solutions for reasoning about time-varying process scenarios are scarce. This paper proposes a method for dynamic context computation concerning spatial and attributive information. Context is interpreted as a body of information dynamically created by a pattern of entities and relationships over a history of situations. Time is conceived as a causative force capable of changing situations and acting on people and objects. The invariant and variant spatial information is captured by a two-dimensional spatial feature representation matrix (SFR-matrix). The time-dependent changes in the context information are computed based on a dynamic context information (DCI) management hyper-matrix. This humble but powerful representation lends itself to a quasi-real time computing and is able to provide information about foreseeable happenings over multiple situations. Based on this, the reasoning mechanism proposed in this paper is able to provide informative instructions for users who needed to be informed in a dynamically changing situation. This paper uses the practical case of evacuation of a building in fire both as an explorative case for conceptualization of the functionality of the computational mechanism and as a demonstrative and testing application. Our intention is to use the dynamic context computation mechanism as a kernel component of a reasoning platform for informing cyber-physical systems (I-CPSs). Our future research will address the issue of context information management for multiple interrelated spaces.@en