A small-arm simulator was prototyped for military simulation training to get shooters acquainted with a geo-located scenario at Royal Thai Army Infantry. GIS was employed to model the entire scenario in an environment and later on simulated in a manner that resembled reality.
To prepare troops for combat-ready capabilities, it is of essence to expose them to a firearms mechanism and to help them acquire needed skills to handle the weapon and to ultimately make a to-shoot-or-not-to-shoot decision. Traditional firearm training costs irreversible ammunitions and expensive maintenance, and most importantly poses live firing risks. On the other hand, small-arm simulator is safer, has lower cost of operation, and more sufficient.
The proposed training system is an ideal combination of training devices, system networking, and application software, which makes up a computer-based trainer, range management software, and an after-action-review tool. The small-arm simulator main systems comprise of shooter or player, sensor or projector and tracking camera, instructor operation and virtual range for scenario screen display. The virtual range system requires ground data collection for geo-specific terrain modelling to establish basic elements of the required feature of this system. Field survey was expected to increase the accuracy and completion of the terrain features to simulated environments, a direct approach to extract reality to geo-specific simulation for training services.
Schematic view of the proposed training system
Components of the small-arm simulator
The virtual range system for generating simulated environments form a critical part of small-arm simulator components that include scenario processing or Unity3D, laser pointer detector, windows message or mouse event, image filtering or blob detection, and video capturing.
Scenario processing was carried out within Game Engine called Unity3D to process and display 3D data in a page display manner that stores users’ interaction in real-time and processes displays user dialogs. Laser pointer tracking detects laser points upon the screen for use in further processing. Scenario processing was sent to the window message through a mouse click event with x, y coordinates incident upon the screen. Laser detection points falling upon the screen were read essentially to convey each image from video frame capture module processing called the blob detection. Finally, the reading frame of the video camera was used to read the screen for mapping locations.
Data collection for GIS-based virtual range
Ground data collection was carried out using the following devices:
- Notebook computer with MapWindow GIS installed was assigned as a ground survey image manager used for the collection of photos and video from and to other devices. It was also a platform for the display of maps and other geospatial data.
- Mobile devices were used as a device for taking photos and immediate data storage. The smart phones HTC one V were all GPS-embedded due to the requirement to collect high precision of locations on the terrain of the photographed 3D objects. To adopt this method, it was realised that spatial accuracy in both horizontal and vertical is completely dependent upon this specific feature of the smart phone used.
- GPS Receiver was another device to collect the information of the position and height of 3D features on the terrain and the predefined route automatically. These along-the-route locations were recorded for later use as a reference of the data collected by the mobile devices and the simulated environments while exploring the route planned.
- GPS mounted Camera was used for shooting height-profile locations and panoramic views of study areas for a comparison between the virtual and real environments.
- Wifi Router was used for a wireless network to back up the data transmission from mobile devices to the image manager as the data transmission for mobile devices could not be done through the mobile phone network.
- Laser Range Finder was used to measure the distance of less than 30 meters from an observer. In densely forested terrain, this equipment was found to be impractical to transmit and receive signals.
Simulated Environment: an interface of GIS and Game Engine
Necessary datasets for creating the virtual world include Digital Elevation Model (DEM), 3D models and imagery data for digitising GIS coordinates and attribute data. GIS and image processing technology has been used to generate real-time simulator databases for regions in most parts of the world. This process consisted of two main parts: 1) Creating a bare 3D terrain from Digital Elevation Model and 2) Placing 3D objects onto the terrain based on their geographical locations.
This process started from digitising imagery data using MapWindow GIS package to identify objects in the scene and to specify attributes for each object. These attributes include object type, latitude and longitude coordinates, scale, rotating angle, and other object-specific features. Another component needed for this process was a collection of 3D models for representing real world objects such as trees, buildings, roads, utility poles etc. in the virtual world. These models had to be imported to Unity as assets so that they were available to the script. After object attributes and 3D assets were ready, the next step was to execute the Unity script. This script got the necessary information from the object attributes such as which 3D model to use for each object, how to scale and rotate the model, and the coordinates for placing the model onto the terrain. The final output was a 3D scene with 3D objects of appropriate types placed in the correct geographical locations.
Technological demonstration: constitutive familiarity and constructive recognition
Thorough terrain analysis is a very important part in the training services of the Army Infantry. It is essential that trainees are familiar with the environment in which they are immersed during training. To create awareness among commanding officers, junior and senior-level practitioners in units of the RTA infantry across the country, the small-arm simulator which was designed for required mobility was transported thereto and demonstrated therein. This technological demonstration was twofold, i.e. to gather user requirements for system enhancement and to collect field survey and ground truth data.
DTI researchers and an army trainee during demonstration
The scenario that took place on the simulated environment of five selected sites were rendered through the scenario processing before 80 commanding officers, junior and senior-level practitioners of the Army Infantry units at the time of technological demonstration.
Following the visualisation of the rendered scenes, they were asked whether they were familiar with the selected sites. With help from an image map showing the whole country, more than 80% of the viewers managed to match the 3D rendered scenes with their real locations. When this group was asked randomly and personally, they responded with no surprise that they knew the real place and at least 60% of them have been to the sites on mission before. Twenty six officers were local of those places, then, familiar almost instantly with the scene.
The simulated environment presented to viewers
There were 38 subjects who managed to identify predominant objects such as big and high trees assumed to be a land mark of the place geographically placed on terrain exactly where they were. Although the script was written to construct the necessary GIS attributes for objects such as the 3D models being selected, the scale and rotation being computed, and the coordinates for models upon the terrain being digitised, the low photorealism that caught viewers’ eye calls for an interface of GIS and Game Engine that magically plays with human eyes to visualise geo-specific scenarios.
Strategic collaboration of research and development
In addition, the tactically detailed activities of this project involved geospatial experts and expertise from various institutions across the country, which drive strategic collaboration of research and development. The Department of Land Development of the Ministry of Natural Resources and Environment whose GIS data include orthophotos, landuse and digital elevation models is considered complete, accurate and timely was consulted and the agreement was reached in form of Memorandum of Understanding (MoU). Geo-Informatics and Space Technology Development Agency (GISTDA) provided expertise through high-resolution satellite images covering the selected sites.
Experts from academic sector are another participative importance of the collaboration since they work in or nearby the vicinity of the selected sites, making the collection, updates and ground truth of the geospatial data practically accessible, inexpensively viable and locally extensive. Private sectors are the major source of hardware and equipment provider which is the beginning of self-reliance for sustainable defense and security. The agreement and collaboration will grow as more projects under the simulation and training Master plan are launched. All these issues truly reflect the Army requirements of geo-specific scenarios that are extensive in training areas, intensive in terrain details and comprehensive in training services.
Various experts and expertise for research and development collaboration
The small-arm simulator is an obvious evidence of the proof-of-concept and the means to gather user requirements from the Army point of view. Military officers who have no ideas of what landscape or terrain they are assigned with a particular mission might look like will be within one mouse click or so to the site of virtual training. Thorough terrain analysis that is indispensable to Infantry training courses was reported to provide constitutive familiarity and constructive recognition via 3D terrain visualisation and modelling. This prototype reaffirms the significance of geospatial data and techniques especially 3D modelling to training services of the Army Infantry. Geospatial data does play a crucial role in the extraction of geographic bit to connect geo-specific scenarios to military simulation and training. However, a lot more needs further and intensive scrutiny to render a proper level of photorealism that meets powerful visualisation requirements.