In the last few years, there has been an increasing demand for street-level simulation environments. Whether it be for tanks, UAVs, dismounted soldiers, or helicopters, each type of simulator needs to be looking at the same thing, at the same time, and with the same quality. The need for this type of high-quality, accurate, urban, immersive 3D environment can raise many challenges for those seeking to create them on a large scale.

In our last blog, we described the path Presagis took to be able to automatically generate massive terrains with little human intervention. We outlined the basic steps to achieve terrain, roads, and buildings using either public, commercial, or private data, or a mix of each.

Continuing on the path to creating realistic, immersive large-scale synthetic environments, this article will outline how vegetation is generated and placed, how textures are positioned, and how materials are assigned to surfaces so that the environment is sensor-ready.

Seeing the Forest for the Trees

“Nature is an excellent designer. So creating realistic or accurate tree placements and groupings is not as simple as randomly dropping trees on to a map,” explains Lead Architect Dave Lajoie.

“When trees grow, their shape is a function of their environment. A cluster of trees, for example, will have a taller one casting shade on smaller ones and that affects the natural look of the cluster. So, vegetation shaded by a building will have a different shape or color than vegetation spread across an open park,” he adds.

In VELOCITY, Presagis aims to recreate that realism. Through imagery, point clouds and other data, VELOCITY analyzes the locations of trees as well as their volume and strives to restore the canopy silhouette depicted in the imagery to duplicate the organic placement and volume of vegetation in the actual city.

“Creating and placing vegetation in this way just looks more natural and goes a long way in creating a realistic and immersive environment,” says Lajoie.

Textures & Materials

With road networks in place, accurate buildings erected, and realistic vegetation planted, the task of texturing and materializing the environment is next.

Applying textures – whether on the ground, or on buildings – is a straightforward task if GIS data is available, and good-to-high resolution textures are accessible. What is more challenging is the application of textures in areas with limited GIS data, or no imagery at all. When automatically creating geo-typical – or procedurally generated – areas, the placement of textures on buildings relies on its proximity to the road, which side is facing the road, the size of the building, how it is connected to power lines, and so on. VELOCITY, in this case, leans on Terra Vista’s ability to recognize road attributes in determining not only how and where a texture is placed on to a building, but how the building meshes with the area, and environment in general.

Lajoie recognizes that textures are only half the battle. “Today, there are simulators with sensors that can zoom in on a very close area that will require a great deal of detail: roads, streets, building details, etc. So just creating a landscape will not cut it,” he explains. “Many general 3D tools don’t implement the concept of material classification. That is, being able to identify that a piece of a building is made of wood or concrete in order to produce the right visual signature when you are looking at it through a sensor like infrared, thermal, or radar is critical.”

In simulation, material classifications can be have a wide variety of uses including sensors, dynamic environments, destruction, and weather.

For this reason, VELOCITY possesses the ability to define materials – such as glass, wood, metal, etc. – for structures and terrains so they have the proper classification when used in sensor applications. To achieve this, VELOCITY performs additional processing that can identify and assign the proper material types, thus enabling sensor technology – including NVG, infrared, thermal, and radar.

VELOCITY permits users to automate the generation of a fully materialized or classified synthetic environment. In turn, this allows users to simulate under any lighting conditions and be ready for any type of sensor.

Output to Virtually Any Format

Outputting or publishing a synthetic environment is the final, yet most important step in producing a useful database. Presagis understands that not every customer works with the same format, and often, a single customer can use a multitude of formats. VELOCITY is able to support a wide array of open standards (legacy and new), as well as streaming services, while remaining format agnostic. Thus, VELOCITY can be combined to feed other systems that will produce environments for Mapping & Visualization, Intelligence Analysis, Decision Support products, as well as content for gaming engines.

Whether the synthetic environment will be used on similar systems, connected systems, complimentary systems, or completely separate ones, it is often necessary that they all be running the same environment. LVC simulation/training, joint training, multi-domain simulations as well as a mix of new and legacy systems all require accurate correlation to deliver consistency and fair fight scenarios across systems. To that end, VELOCITY is able to output to the most common simulation formats such as OpenFlight, OneSaf, VBS3, JSAF, and OGC CDB.

In addition to those formats, Presagis understands there is a convergence between the technologies traditionally used to create environments and simulations, and those that create games. Hence, VELOCITY leverages the power of Terra Vista to output multiple formats from the same source data. So, once your environment is built, you can output terrains to a very wide variety of formats including VBS3, OGC CDB, OpenFlight, and even FBX to be used in gaming engines.

This flexibility isn’t just to accommodate the different formats used by different simulators, it also opens many more doors when it comes to file exchange whether it be for interoperability, or digital content creation.

Summary

“90% of the world’s data had been created in the last two years…” -IBM

And it’s only growing.

Over the course of our last two blogs on building large-scale environments, we described the way in which Presagis has approached this challenge using VELOCITY. From military training to smart cities, organizations facing the crush of big data, or those seeking to leverage this explosion of information need a way to manage, fuse, and automate their virtual environments.

In building VELOCITY, Presagis has developed a solution that is:

• a powerful way to generate large-scale 3D environments in an automated manner
• scalable so that processing can be distributed (even on the cloud) and can be implemented within any infrastructure
• the most flexible manner of using the widest possible range of inputs
• a way to ensure data is correlated across systems
• built to fuse multiple data sources – from GIS data and point clouds to imagery and sensors
• designed for frequent, repeated data updates

This is by no means the only way to build vast synthetic environments, but, in our experience, this solution offers a large number of benefits for any organization that has massive amounts of data, requires frequent database updates, and requires large-scale synthetic environments. The world is generating 2.5 quintillion bytes of data per day, and with the exploding growth of the Internet of Things, and autonomous vehicles, this number may be conservative. The data tsunami is coming… are you ready?