Abstracts Track 2023

Area 1 - Simulation Technologies, Tools and Platforms

Nr: 15

Model-Driven Data Storage Using Dynamic Object-Relational Mapping


Piet Van Der Paelt, Ben Lauwens and Beat Signer

Abstract: Many programming languages offer discrete-event simulation frameworks, each of them varying in completeness of offered features and ease of use. This research focusses on the SimJulia package (Lauwens et al.), which is an open source discrete-event simulation framework, available within the Julia programming language ecosystem. In the domain of Simulation Tools and Platforms, we present a novel transparent data persistence architecture as an extension of SimJulia. We integrated the PostgresORM (Laugier et al.) package into ResumableFunctions (Lauwens et al.), on which SimJulia depends heavily, by using Julia's language features (metaprogramming). As such, we were able to remove the dependency on a user's knowledge on architectures for persistence. Our contribution aims to improve the usability of the SimJulia package, whilst demonstrating the power of Julia's macro expansion to move towards a dynamic object-relational mapping (ORM) configuration. This intermediary result is part of a broader research project in which we aim to realise a web-based architecture for modelling, simulation and visualisation in support of research and education simulation requirements. SimJulia is offered through Julia's general package repository and currently, must be used within a development environment of choice. As such, it is a widely available open source technology. However, the current implementation lacks functionality to transparently store state variables of a simulation. The exploitation of the data generated by a SimJulia simulation depends entirely on the user's knowledge on technology for persistence and how to interact with it from within the simulation model. To mitigate this shortcoming, we extended both the SimJulia and the ResumableFunctions packages by implementing a transparent probing and data persistence architecture, employing Julia's metaprogramming support. As such, the user is liberated from technological and infrastructural concerns to realise data persistence. Furthermore, the dynamic aspect of our implementation reduces required configuration with regard to the persistence aspect to a strict minimum. Aside the aforementioned Julia packages and language features, the persistence infrastructure is realised through the PostgreSQL Relational Database Management Systems (RDBMS). As a result of our implementation, the end-users have the possibility to store data which characterises the evolution of a running simulation transparently. Aside the requirement of having a RDBMS instance running, there is no additional configuration burden. The data is available in the database and can be exploited using any technology of choice. Apart from the gain in functionality, there is also a technological gain which is situated at the usage of metaprogramming to divert from a static ORM configuration as often seen in web applications, to a transparent and dynamic configuration. As a part of our broader research project, the current proof-of-concept implementation also demonstrates the exploitation of the stored data in a visualisation which shows the tables in a web interface. The latter is realised using the VueJS package. Through the intermediary of an abstraction layer which employs the dynamic ORM, we were able to provide VueJS with the required view on the data. We are currently considering the possibility to externalise the data based on a REST API for the presented architecture.

Nr: 78

Development of Paralle Microwave Analysis OSS: ADVENTURE_Fullwave


Amane Takei

Abstract: In this presentation, a parallel microwave analysis open source software:OSS based on an iterative domain decomposition method is explained that is named ADVEN-TURE_Fullwave (https://adventure.sys.t.u-tokyo.ac.jp/). A stationary vector wave equation for the high-frequency electromagnetic field analyses is solved taking an electric field as an unknown function. Then, to solve subdomain problems by the direct method, the direct method based on the LDL^T decomposition method is introduced in subdomains. The simplified Berenger’s PML is introduced which these eight corners are given the average value of all PML’s layers.

Nr: 79

Fast Mesh Generation Techniques for Huge Scale Microwave Analysis


Hiroshi Kawai

Abstract: We have been conducting a huge scale analysis for the full-wave electromagnetic field using the finite element method on various kinds of massively parallel supercomputers, such as K Computer and Fugaku in Japan. Through the whole simulation task including its pre- and post-processing, however, now the mesh generation step becomes the clear bottleneck. Currently, we are developing a fast and robust tetrahedral mesh generation technique based on the quad-tree/octree method. Detail of the mesh generation algorithm will be explained and some examples of huge scale mesh generation will be demonstrated in the presentation.

Area 2 - Application Domains

Nr: 72

Improvement in Ergonomics for Firefighters Using Digital Human Modeling


Yuan Xing Xu, Rensheng Deng, Noor Asrik Masahlan, Nur Irwan Abu Bakar, Syed Yazid Syed Nasiruddin, Md Erwandy Md Ali, Saravana Kumarasamy, Meng Fai Ying and Alan Chow

Abstract: Background Firefighters often carry heavy backpacks during their operations and one such backpack carried by them is the compressed air foam (CAF) backpack which weighs 20kg and is stored in the rear cabin of the light fire attack vehicle (LF5G*). CAF backpack is designed for light firefighting and is more effective than normal fire extinguishers. Firefighters must lean forward into the cabin and drag the backpack along an upwards slope to the edge of the cabin before bending down to don it, which puts them in an awkward position. An assessment of these actions using digital human modeling (DHM) has shown that firefighters experience high risk of musculoskeletal disorders. Approach The addition of a horizontal sliding platform to the rear cabin of LF5G was proposed to remove the need for firefighters to lean forward and bend down during the retrieval and donning of the backpack. The extent of improvement brought about by the modification was evaluated with DHM. The scenario of a firefighter retrieving the backpack is reconstructed in DHM, with the rear cabin and backpack represented by rectangular blocks. The dimensions of the blocks mirror the exact dimension of the rear cabin and backpack. 2 different scenarios were constructed to represent pre and post modification of the rear cabin and DHM was then used to simulate the motion of a firefighter retrieving the backpack in both scenarios. DHM comes with ergonomics analysis tools and metrics that were collected including the Ovako Working Posture Analysis (OWAS), Rapid Upper Limb Assessment (RULA), Lower Back Compression Analysis (LBCA) and Static Strength Prediction (SSP). These metrics provide a quantitative ergonomic assessment of the changes proposed. Result Through the simulation, it was found that the modification resulted in a significant improvement in 3 out of 4 metrics and thus reduced the risk of musculoskeletal disorders experienced by firefighters. OWAS remain unchanged, reflecting insignificant improvement in working posture. The grand score of RULA was reduced from 6 to 4, indicating a reduction in injury risk of upper limbs. The lower back compression force (on L4-L5 portion of the spine) was reduced by 24% from 1973N to 1501N upon the modification. SSP indicated that a vast majority of the population will be able to handle the load after the modification. Conclusion The simulation provided promising results and justification for the modification. Currently, 4 vehicles have been modified as part of a trial to obtain user feedbacks with the aim of implementing the modification across all LF5G vehicles in the future. *Picture of the LF5G can be found here: https://images.app.goo.gl/jzBKedhyh9C3vzNk8.