Producing Synthetic Dataset for Human Fall Detection in AR/VR Environments

Human poses and the behaviour estimation for different activities in (virtual reality/augmented reality) VR/AR could have numerous beneficial applications. Human fall monitoring is especially important for elderly people and for non-typical activities with VR/AR applications. There are a lot of different approaches to improving the fidelity of fall monitoring systems through the use of novel sensors and deep learning architectures; however, there is still a lack of detail and diverse datasets for training deep learning fall detectors using monocular images. The issues with synthetic data generation based on digital human simulation were implemented and examined using the Unreal Engine. The proposed pipeline provides automatic “playback” of various scenarios for digital human behaviour simulation, and the result of a proposed modular pipeline for synthetic data generation of digital human interaction with the 3D environments is demonstrated in this paper. We used the generated synthetic data to train the Mask R-CNN-based segmentation of the falling person interaction area. It is shown that, by training the model with simulation data, it is possible to recognize a falling person with an accuracy of 97.6% and classify the type of person’s interaction impact. The proposed approach also allows for covering a variety of scenarios that can have a positive effect at a deep learning training stage in other human action estimation tasks in an VR/AR environment.

Digital Twin Based on 3D Visualization and Computer Image Recognition Technology in Coal Preparation Plant

The digital twin coal preparation plant is a potentially effective way to realize the intelligent interconnection and interactive integration of the manufacturing physical world and the information world. Aiming at the difficulty in predicting and maintaining the state of the shearer in a harsh working environment, combined with the high-fidelity behaviour simulation characteristics of the digital twin and the powerful data mining capabilities of deep learning, a coal shearer health prediction driven by the integration of the digital twin and deep learning is proposed. Method. The article builds an information management and integration platform composed of a real-time database and a comprehensive information platform, and realizes the unified management of data integration and application systems.

ABWiSE v1.0: toward an agent-based approach to simulating wildfire spread

Wildfires are a complex phenomenon emerging from interactions between air, heat, and vegetation, and while they are an important component of many ecosystems’ dynamics, they pose great danger to those ecosystems, as well as human life and property. Wildfire simulation models are an important research tool that help further our understanding of fire behaviour and can allow experimentation without recourse to live fires. Current fire simulation models fit into two general categories: empirical models and physical models. We present a new modelling approach that uses agent-based modelling to combine the complexity possible with physical models with the ease of computation of empirical models. Our model represents the fire front as a set of moving agents that respond to, and interact with, vegetation, wind, and terrain. We calibrate the model using two simulated fires and one real fire and validate the model against another real fire and the interim behaviour of the real calibration fire. Our model successfully replicates these fires, with a figure of merit on par with simulations by the Prometheus simulation model. Our model is a stepping-stone in using agent-based modelling for fire behaviour simulation, as we demonstrate the ability of agent-based modelling to replicate fire behaviour through emergence alone.

Reframing Wildfire Simulations for Understanding Complex Human–Landscape Interactions in Cross-Cultural Contexts: A Case Study from Northern Australia

An increase in the frequency of severe fire events, as well as a growing interest in wildfire mitigation strategies, has created a demand for skilled managers of landscape fire and a better community understanding of fire behaviour. While on-ground experience is essential, there is potential to substantially enhance training and community engagement with explanatory simulations. Through this work, we explore landscape fire behaviour as a complex system where understanding key behaviour characteristics is often more important and achievable than prediction. It is argued that this approach has particular value in Northern Australia, where fires burn across vast and sparsely inhabited landscapes that are largely under Indigenous ownership. Land and fire management in such complex cross-cultural contexts requires combining traditional and local knowledge with science and technology to achieve the best outcomes. We describe the workings of the model, a stochastic cellular automata fire behaviour simulation, developed through a participatory modelling approach for Northern Australia; the outputs generated; and a range of operational applications. We found that simulation assisted training and engagement through the development of an understanding of fire dynamics through visualisation, underpinned by landscape data sets, and engaging a culturally diverse set of land managers in discussions of fire management. We conclude that there is scope for a broader use of explanatory fire simulations to support development of shared understandings of fire management objectives.

ABWiSE v1.0: Toward and Agent-Based Approach to Simulating Wildfire Spread

Wildfires are a complex phenomenon emerging from interactions between air, heat, and vegetation, and while they are an important component of many ecosystems’ dynamics, they pose great danger to those ecosystems, and human life and property. Wildfire simulation models are an important research tool that help further our understanding of fire behaviour and can allow experimentation without recourse to live fires. Current fire simulation models fit into two general categories: empirical models and physical models. We present a new modelling approach that uses agent-based modelling to combine the complexity found in physical models with the ease of computation of empirical models. Our model represents the fire front as a set of moving agents that respond to, and interact with, vegetation, wind, and terrain. We calibrate the model using two simulated fires and one real fire, and validate the model against another real fire and the interim behaviour of the real calibration fire. Our model successfully replicates these fires, with a Figure of Merit on par with simulations by the Prometheus simulation model. Our model is a stepping-stone in using agent-based modelling for fire behaviour simulation, as we demonstrate the ability of agent-based modelling to replicate fire behaviour through emergence alone.

Wheel flat detection and severity classification using deep learning techniques

Wheel flats are one of the most common types of defect found in railway systems. Wheel flats can result in decreasing passenger comfort and noise if they are slight, or serious incidents such as derailment if they are severe. With the increasing demand for railway transport, the speed and weight of rolling stock tend to increase, which results in relatively rapid deterioration. The occurrence of wheel flats is also affected by this increasing demand. To perform preventative maintenance for wheel flats, to keep wheelsets in a proper condition and to minimise maintenance costs, the ability to detect and classify wheel flats is required. This study aims to apply deep learning techniques to detect wheel flats and classify wheel flat severity. The deep learning techniques used in the study are a deep neural network (DNN), a convolutional neural network (CNN) and a recurrent neural network (RNN). 1608 samples, simulated using D-Track, a dynamic behaviour simulation software package, are used to develop machine learning models. Three different aspects of the models are evaluated, namely overall accuracy, the ability to detect wheel flats and the ability to classify wheel flat severity. The results from the study show the DNN has the highest overall accuracy of 96%. In addition, the DNN can be used to detect wheel flats with nearly 100% accuracy. The CNN performs better than the RNN in terms of overall accuracy and wheel flat detection. However, the RNN performs better than the CNN in wheel flat severity classification. Overall, the DNN offers the best approach for detecting wheel flats and classifying their severity.

Forest fuel assessment by LiDAR data. A case study in NE Italy

<p>Fuel management is a crucial action to maintain wildland fires under the threshold of manageability; hence, in order to allocate resources in the best way, wildland fuel mapping is regarded as a necessary tool by land managers. Several studies have used Aerial Laser Scanner (ALS) data to estimate forest fuels characteristics at plot level, but few have extended such estimates at a zonal level.</p><p>In the context of the EU Interreg Project CROSSIT SAFER, a test of the possibilities of ALS data to predict fuels attributes has been performed in three different areas: an alpine basin, a coastal wildland-urban interface and a karstic highland. Eighteen sampling plots have been laid out over 6 forest categories, with a special focus on <em>Pinus nigra</em> J. F. Arnold artificial forests. Low density (average 4 points/m<sup>2</sup>) discrete return LiDAR data has been analysed with FUSION, a free point cloud analysis software tailored to forestry purposes; field and remote sensing data have been connected with simple statistical modelling and results have been spatialised over the case study areas to provide wall-to-wall inputs for FLAMMAP fire behaviour simulation software.</p><p>Resulting maps can be of relevance for land managers to better highlight the most vulnerable or fire prone areas at a mesoscale administrative level. Limitations and room for improvement are pointed out, in the view that land management should keep updated with the latest technology available.</p>