In the last few years, the deep learning (DL) computing paradigm has been deemed the Gold Standard in the machine learning (ML) community. Moreover, it has gradually become the most widely used computational approach in the field of ML, thus achieving outstanding results on several complex cognitive tasks, matching or even beating those provided by human performance. One of the benefits of DL is the ability to learn massive amounts of data. The DL field has grown fast in the last few years and it has been extensively used to successfully address a wide range of traditional applications. More importantly, DL has outperformed well-known ML techniques in many domains, e.g., cybersecurity, natural language processing, bioinformatics, robotics and control, and medical information processing, among many others. Despite it has been contributed several works reviewing the State-of-the-Art on DL, all of them only tackled one aspect of the DL, which leads to an overall lack of knowledge about it. Therefore, in this contribution, we propose using a more holistic approach in order to provide a more suitable starting point from which to develop a full understanding of DL. Specifically, this review attempts to provide a more comprehensive survey of the most important aspects of DL and including those enhancements recently added to the field. In particular, this paper outlines the importance of DL, presents the types of DL techniques and networks. It then presents convolutional neural networks (CNNs) which the most utilized DL network type and describes the development of CNNs architectures together with their main features, e.g., starting with the AlexNet network and closing with the High-Resolution network (HR.Net). Finally, we further present the challenges and suggested solutions to help researchers understand the existing research gaps. It is followed by a list of the major DL applications. Computational tools including FPGA, GPU, and CPU are summarized along with a description of their influence on DL. The paper ends with the evolution matrix, benchmark datasets, and summary and conclusion.

/pdf/review-of-deep-learning-concepts-cnn-architectures-hpbk62qzc2.pdf

Review of deep learning: concepts, CNN architectures, challenges, applications, future directions

安全装置工業事故豫防法 = Industrial accident prevention

The purpose of this paper is to explore how rising technologies from Industry 4.0 can be integrated with circular economy (CE) practices to establish a business model that reuses and recycles wasted material such as scrap metal or e-waste.,The qualitative research method was deployed in three stages. Stage 1 was a literature review of concepts, successful factors and barriers related to the transition towards a CE along with sustainable supply chain management, smart production systems and additive manufacturing (AM). Stage 2 comprised a conceptual framework to integrate and evaluate the synergistic potential among these concepts. Finally, stage 3 validated the proposed model by collecting rich qualitative data based on semi-structured interviews with managers, researchers and professors of operations management to gather insightful and relevant information.,The outcome of the study is the recommendation of a circular model to reuse scrap electronic devices, integrating web technologies, reverse logistics and AM to support CE practices. Results suggest a positive influence from improving business sustainability by reinserting waste into the supply chain to manufacture products on demand.,The impact of reusing wasted materials to manufacture new products is relevant to minimising resource consumption and negative environmental impacts. Furthermore, it avoids hazardous materials ending up in landfills or in the oceans, seriously threatening life in ecosystems. In addition, reuse of wasted material enables the development of local business networks that generate jobs and improve economic performance.,First, the impact of reusing materials to manufacture new products minimises resource consumption and negative environmental impacts. The circular model also encourages keeping hazardous materials that seriously threaten life in ecosystems out of landfills and oceans. For this study, it was found that most urban waste is plastic and cast iron, leaving room for improvement in increasing recycling of scrap metal and similar materials. Second, the circular business model promotes a culture of reusing and recycling and motivates the development of collection and processing techniques for urban waste through the use of three-dimensional (3D) printing technologies and Industry 4.0. In this way, the involved stakeholders are focused on the technical parts of recycling and can be better dedicated to research, development and innovation because many of the processes will be automated.,The purpose of this study was to explore how Industry 4.0 technologies are integrated with CE practices. This allows for the proposal of a circular business model for recycling waste and delivering new products, significantly reducing resource consumption and optimising natural resources. In a first stage, the circular business model can be used to recycle electronic scrap, with the proposed integration of web technologies, reverse logistics and AM as a technological platform to support the model. These have several environmental, sociotechnical and economic implications for society.,The sociotechnical aspects are directly impacted by the circular smart production system (CSPS) management model, since it creates a new culture of reuse and recycling techniques for urban waste using 3D printing technologies, as well as Industry 4.0 concepts to increase production on demand and automate manufacturing processes. The tendency of the CSPS model is to contribute to deployment CE in the manufacture of new products or parts with AM approaches, generating a new path of supply and demand for society.

Exploring Industry 4.0 technologies to enable circular economy practices in a manufacturing context: A business model proposal

This paper concludes that 4D models are a useful alternative to project scheduling tools like CPM networks and bar charts. They enable more people to understand a schedule quickly and identify potential problems. By developing a 4D model for a commercial construction project, we were able to detect the incompleteness of the original schedule, find inconsistencies in the level of detail among the schedule activities, and discover an impossible schedule sequence. We were also able to anticipate potential time-space conflicts and accessibility problems. The results of the case study show that 4D models are effective in evaluating the executability of a construction schedule. The case study also highlighted the need for improvements to 4D tools. 4D tools should include bar charts, component lists, and annotation tools in their graphical user interface. Automating schedule data preparation and 4D model generation in the design stages of a project can expedite 4D model development and use. Users need to be able to generate 4D models at multiple levels of detail and generate and evaluate alternative scenarios rapidly.

Hanbin Luo

Papers

A deep hybrid learning model to detect unsafe behavior: Integrating convolution neural networks and long short-term memory

Detecting non-hardhat-use by a deep learning method from far-field surveillance videos

A BIM-based construction quality management model and its applications

Falls from heights: A computer vision-based approach for safety harness detection

Automated detection of workers and heavy equipment on construction sites: A convolutional neural network approach