Towards Accuracy Enhancement of Age Group Classification Using Generative Adversarial Networks

Age estimation models can be employed in many applications, including soft biometrics, content access control, targeted advertising, and many more. However, as some facial images are taken in unrestrained conditions, the quality relegates, which results in the loss of several essential ageing features. This study investigates how introducing a new layer of data processing based on a super-resolution generative adversarial network (SRGAN) model can influence the accuracy of age estimation by enhancing the quality of both the training and testing samples. Additionally, we introduce a novel convolutional neural network (CNN) classifier to distinguish between several age classes. We train one of our classifiers on a reconstructed version of the original dataset and compare its performance with an identical classifier trained on the original version of the same dataset. Our findings reveal that the classifier which trains on the reconstructed dataset produces better classification accuracy, opening the door for more research into building data-centric machine learning systems.

One-of-a-kind Production in Cyber-Physical Production Systems Considering Machine Failures

Industry 4.0 proposes the incorporation of information technologies at all levels of the production process. By incorporating these technologies, Industry 4.0 provides new tools for production planning processes, allowing to address problems in an innovative and efficient manner. From these technologies and tools, it is that in this work a One-of-a-Kind Production (OKP) process is approached, where the products tend to be highly customized. OKP implies working with a very large variability within production, demanding very efficient planning systems. For this, a planning model based on CONWIP-type strategies was proposed, which seeks to level the production of a shop floor configured in the form of a job shop. Even more, for having a more realistic shop-floor representation, machine failures have been included in the model. In turn, different dispatching rules were proposed to study the performance and analyze the behaviour of the system. From the results obtained, it is observed that, when the production demand is very exigent in relation with the capacity of the system, the dispatching rules that analyze the workload generated by each job tend to perform better. However, when the demand on the capacity of the production system is less intense, the rules associated with due dates are the ones that obtain the best results.

A Prioritized Decision Making Method for Reliability Allocation: in Optimistic and Pessimistic View

Reliability allocation is a very important problem during early design and development phases of a system. There are several reliability allocation techniques which are used to achieve the target reliability. The feasibility of objectives (FOO) technique is one of them that is widely used to perform system reliability allocation. But this technique has two fundamental shortcomings. The first is the measurement scale and the second is that it does not consider the order weight of the reliability allocation factors. The prioritization of the factors is also an important topic in decision making. Practically, all factors in multi-criteria decision making (MCDM) are not in the same priority level. Hence, in decision making situation, it is usual for decision makers to consider different priority factors. So, considering the prioritization of the factors, a reliability allocation method is proposed here to overcome the shortcomings of the FOO technique. Also, a case study on reliability allocation in airborne radar system is considered here to verify the efficiency of the proposed approach. Finally, the results are calculated in different optimistic and pessimistic view point and compared with the FOO technique. This comparison exhibits the advantages and supremacy of the proposed approach.

Effectiveness of different requirements checklists for novice designers

Working under constrained conditions can boost or kill creativity, depending on the nature of the constraints (organizational, personal or task-related). However, a design process without clearly identified constraints, which set the project objectives, could lead to inefficiencies and unfruitful iterations. Some of the most acknowledged procedures to support requirement definition are focused on the use of specific checklists. However, notwithstanding the importance of the task, little attention was dedicated to the verification of the effectiveness of these tools. In such a context, the paper presents an investigation aimed at assessing the performance of three checklists that exploit different strategies to elicit requirements. To that purpose, a sample of fifty engineering students was asked to use the checklists to define the requirements for a specific design case. The outcomes of the experiment were assessed according to well-acknowledged effectiveness metrics, i.e. quantity, operationality, validity, non-redundancy, and completeness. The result of the assessment highlights that checklists based on more general questions or abstract stimuli can better support novice designers in making explicit internally felt design constraints that can potentially lead to more innovative design.

Identification and Assessment of Obsolescence in the Early Stages of System Design

Obsolescence is the fact that an entity (physical or logical) is becoming outdated or no longer possesses the required level of performance. The objectives of this article are twofold. First, it is intended to contribute to the understanding of obsolescence propagation. Secondly, two supporting approaches for the Identification and Assessment phases are proposed: the House of Obsolescence and the System Obsolescence Criticality Analysis. The former allows the mapping of obsolescence propagation via dependencies, whether imposed changes are desired or imposed, by external actors to the system architecture. Whereas, the objective of the latter is to assign an obsolescence criticality index to the identified risks in order to prioritize them for solution or mitigation determination during the analysis phase. The tools make extensive use of the modeled system knowledge through the application of Systems Engineering. The application of these approaches is presented through an illustrative study.

Connecting Building Design with the Digital Factory by Design Languages to Explore Different Solutions

In engineering, design decisions in one domain exhibit multiple consequences in other domains. These consequences result from the often more or less hidden coupling between the different design domains. In order to examine these consequences, models need to be created. In practice, this is challenging due to the exchange of data between different engineering domains, since different software applications are often used and the effort involved with manual model creation. In this paper, we explore the use of graph-based design languages in a Model-Based Systems Engineering (MBSE) approach to link the digital factory with building design. We also show that the use of a common formal representation based on the Unified Modeling Language (UML) supports the interoperability between the two domains. Finally, we demonstrate how the engineering knowledge for the preliminary design of a factory building can be formally described using graph-based design languages and how the production line of the digital factory can then be used as an input to automatically create valid preliminary designs for the factory building.1

Remote Run-Time Failure Detection and Recovery Control For Quadcopters

We propose an adaptive run-time failure recovery control system for quadcopter drones, based on remote real-time processing of measurement data streams. Particularly, the measured RPM values of the quadcopter motors are transmitted to a remote machine which hosts failure detection algorithms and performs recovery procedure. The proposed control system consists of three distinct parts: (1) A set of computationally simple PID controllers locally onboard the drone, (2) a set of computationally more demanding remotely hosted algorithms for real-time drone state detection, and (3) a digital twin co-execution software platform — the ModelConductor-eXtended — for two-way signal data exchange between the former two. The local on-board control system is responsible for maneuvering the drone in all conditions: path tracking under normal operation and safe landing in a failure state. The remote control system, on the other hand, is responsible for detecting the state of the drone and communicating the corresponding control commands and controller parameters to the drone in real time. The proposed control system concept is demonstrated via simulations in which a drone is represented by the widely studied Quad-Sim six degrees-of-freedom Simulink model. Results show that the trained failure detection binary classifier achieves a high level of performance with F1-score of 96.03%. Additionally, time analysis shows that the proposed remote control system, with average execution time of 0.49 milliseconds and total latency of 6.92 milliseconds in two-way data communication link, meets the real-time constraints of the problem. The potential practical applications for the presented approach are in drone operation in complex environments such as factories (indoor) or forests (outdoor).

Research Strategy in 4D Printing: Disruptive vs Incremental?

The paper aims at presenting 4D printing as a research-intensive technology from a critical external perspective. It provides a comprehensive discussion on the possible future of this emerging domain and also highlights weaknesses and strengths of applying a disruptive or incremental research strategy. Most scientific research efforts in 4D printing contribute to developing the spectrum of possible changes by investigating stimulus/smart materials combinations with additive manufacturing technologies. Although the current results are spectacular, the performances are still far from the basic requirements expected in the industry. The paper highlights the current limitations and trends towards incremental research strategies and argues in favor of risk-taking and the disruptive nature of research to make leaps that benefit society. Even if transgressive promises are associated with this technology with high growth potential in academic research, where creativity is involved and related invention derived, targeted applications are far from being achieved leading to a risk of the slow death of the field and unsatisfactory innovation. Based on this assessment, it appears that close fields in a situation of possible disciplinary porosity can – with a little openness and some creativity – move away from the current highly self-centered work to try to rekindle 4D printing, provided that risk-taking in interdisciplinary research is better supported. If creativity and interdisciplinary project management for innovation are to be promoted, the organizational context must be conducive to risk-taking for this redeployment.

Environment-Based Life Cycle Decomposition (eLCD): Adaptation of EBD to Sustainable Design

As sustainability becomes increasingly important, product design is taking a proactive role in producing products that are both useful and sustainable. This paper introduces and discusses a tool named Environment-based life cycle decomposition (eLCD) to adapt the Environment-based Design (EBD) methodology to sustainable design. The eLCD brings to EBD three major features: 1) a holistic environment structure for sustainable conceptual design, 2) an effective and efficient tool for collecting information for sustainability decision-making, and 3) an analysis tool that takes sustainability as an integral part of the design rather than as a burden. The environment of a product is everything except the product itself, which can be defined in three dimensions, namely, environment types, life cycle events, and life cycle time. The environment types are designated as natural, built (including physical artifact and digital artifact), economic, and social environment. The eLCD provides an effective template for information collection to support the design decision-making process. The effectiveness of eLCD is demonstrated by its application to the upscaling of a wind turbine, where an energy storage system is introduced to make full use of wind energy with the least waste in serving the electricity demand.

Performing Information Extraction and Ontology Development for Mission Engineering Applications

The development and evaluation of Mission Engineering Threads (METs) require an understanding of the operational context for which a system-of-systems (SoS) will be employed as well as an assessment that the performance of a complex SoS is effective and safe. The information describing the design and performance parameters of the systems within the SoS is distributed in many different physical locations, and is represented in a variety of formats, both structured and unstructured. Because of this dynamic on the structuring of the data to include differing ontology frameworks, it is necessary to develop a framework and toolset to handle the automated extraction of information from disparate information sources. In addition, this extracted information needs to be categorized properly into defined data types as represented in the specific MET to correctly capture the appropriate context of the mission scenario. Semantic technique solutions will be researched, analysed, and applied as a means to infer new facts from existing facts and data. These techniques are particularly powerful when the amount of data and/or the relationships and constraints among data are too cumbersome and complex for human understanding and reasoning. Using the characteristics of wine as an example, we present our framework to show how it enables rapid and contextually relevant extraction, and represents complex information in a user-friendly format.