FOSS4G 2023

Chemical incidents, such as accidents at heavy chemical plants or large-scale toxic gas leaks, are difficult to assess accurately because of the large spatial extent of the damage and the rapidly changing scope/level/target of the damage over time. These characteristics also make it hard to conduct experiments to recreate or simulate large-scale chemical incidents in real world. In the case of large-scale chemical accidents or release, post-incident damage assessment is as important as prevention, but spatial ambiguity makes it difficult to assess the extent of damage to victims, and there is little way to identify fake victims from real ones. 

In this 5 year-long study, we aim to combine the results of a chemical diffusion model and the location data of mobile service subscribers on the incident spot over time. For this, FOSS4G based 3D geospatial web service using GeoServer, Postgresql/PostGIS, Cesium, etc. will be developed to assess the level of chemical exposure of each victim and calculate the level of damage based on it. 

In 2022, the first year of the study, we developed a prototype that combines the time-dependent output of the chemical diffusion model with the time-dependent location data of individuals and successfully visualized it in a Web 3D globe. In the coming year, we'll further develop this system into an integrated risk assessment platform for chemical accidents by combining chemical exposure assessment model and damage calculation model.

Because environmental impact assessment(EIA) process is a combination of detailed fields that require a lot of expertise (e.g., noise, air pollution, odor, water pollution, ecological environment, living environment, etc.), despite its long history, the process is still complex and slow, and it is not easy to break away from the document/drawing-centered work process. Since the nature of the environment involves many geographic/spatial context, if it can be assisted with a spatio-temporal system, it can be expected to show very high efficiency compared to the current process.

To verify the feasibility of such a system, we adopted a FOSS4G-based approach and developed a pilot system in this study. Specifically, we used GeoServer and Postgresql/PostGIS for handling and providing data spatially, and Cesium for 3D geospatial based visualization. We focused on the design and implementation of APIs to assemble the sub-processes of EIA, as well as the visualization and UI of the pilot system.

This system demonstrates how the noise propagate during and after the construction in an interactive way. We expect the system will increase the non-expert stakeholder's understanding of noise propagation visually.

Through this presentation, we will discuss our findings implemented in a EIA process centered on the noise, from the first step of applying for approval from the civil/construction operator to the last step of deriving the final evaluation opinion by the noise expert in charge, and provide clues to the future of Digital EIA.

In the future, we believe that the expansion to other EIA media and the smooth implementation of current legal and administrative tasks will make it a system that can be used in the field.

How do you run an aid station in case of a disaster? Scenarios are planned for each city, but there are limitations in applying them to actual aid station operations. In our presentation, we will present a case study on the development and simulation of a aid station management tool using digital twin technology and share various visualization techniques in a 3D city model environment.

The study site is Ulju-gun, a county of about 220,000 people in southern South Korea, with two nuclear power plants operating within a few kilometers of each other. Moving people to shelters to protect them in the event of a disaster such as a radioactive leak is very essential and crucial part of disaster management.

The aid station management tool presented in this presentation leverages ground-truth 3D modeling data of the shelter buildings that will be operational during a disaster to provide facility placement and editing capabilities. This allows relief tents to be automatically placed or edited based on the scenario. It also provides the ability to monitor the overall changes that may occur at the shelter through a dashboard, including real-time victim status, food, beverage, and medical support, supply status, shelter information, and disaster situation information.

The Cesium platform is used to service the data and the Three.js library is used to handle the viewing and placement of 3D model data in glTF format. Other open source implementations include React, Turf.js, Apache ECharts, and GeoServer.

We believe that the findings mentioned in this study provide a good example of how 3D city model-based shelter operations and visualization techniques can be applied to disaster preparedness systems to support effective decision-making and resource allocation.

This talk is about the current state of MilMap and its ongoing development. MilMap is a military geo-portal system widely and successfully used in every sectors of Korean military. The system is now undergoing major change from geo-portal to military digital twin system.

MilMap is developed on top of numerous open source projects such as PostGIS, GeoServer, GeoWebCache, Cesium, OpenLayers, mago3D, OpenGXT. The system provides several functionalities like POI search, geospatial data search, layer control, satellite image search and download, spatial terrain analysis, coordinates reading, and map notes, to the military officers through the intranet. Although the system provides geospatial analytics functions through OGC WPS(Web Processing Service), the current system is basically a web based 3D GIS for data viewing and printing. Thanks to MilMap, military officers can now access the huge amount of geospatial data(maps, imagery, 3D, POI, and others) in their browser without installing additional software.

MilMap is now undergoing major development to be a more customized, automated, and analytical system. The future MilMap will support user data uploading for intelligence sharing, more bespoke battle field analysis and others. In the long run, MilMap is expected to be a cloud based military digital twin system for geospatial intelligence sharing and battle field analysis & simulation.