Threshold Requirements for Simulating Terrain for Ground Vehicles
MetaVR™ customers use our 3D virtual-world creation and visualization products to prototype robotic and crew-based future-combat systems vehicles that use 1-meter elevation data and 30 centimeter imagery for their real-time simulated terrain. This high-resolution real-time 3D terrain has many advantages for real-time simulation purposes. Of equal importance, this high-resolution visualization provides a direct link to real-time 3D ground-based situational awareness (that is, augmented reality) as an embedded system that is comparable to those systems used for unmanned aerial vehicles (UAV) embedded systems.
Simulating navigation in robotic vehicles
Unmanned ground combat vehicles have developed more slowly than unmanned aerial vehicles due to the design challenges one encounters from the shear complexity of navigating earth and manmade surfaces and obstacles. Compounding the design challenge is the fact that the smaller the vehicle, the more complex the task of navigating difficult terrain. During the design of such robotic vehicles, analytical simulation models that are not real-time have traditionally been used to study the interaction of the terrain with the vehicle suspension system. With MetaVR technology, our customers can now use the same elevation data model that is used for the analytic simulation for simulating the vehicle dynamics response to terrain feedback points as is used in the real-time simulation at 60 Hz. In conjunction with the software advancements, now available for prototyping embedded systems are MetaVR Low Profile Rackmount VRSG 1U visual systems, which are similar but smaller in size to those used for the Tactical Unmanned Aerial System (TUAS) ground control stations.
Example of effect of elevation resolution on suspension
As you click through the images below you can see that the suspension system of the small 4-wheeled remotely piloted vehicle reacts more dramatically to the higher resolution elevation data used to create the underlying terrain.
MetaVR model of robotic vehicle. Click the image to cycle through the effect of the elevation data on suspension.
More accurate dynamics feedback due to the higher resolution 1-meter terrain will give you a more accurate prediction of how the real vehicle will behave during actual use. Note that an operator may drive the simulated vehicle over terrain at 4-meter resolution that if it were simulated at 1-meter, the vehicle might actually tip over by driving on a steep slope. As the wheel base is smaller for these prototype robotic future combat systems than the Stryker vehicle shown below, it is affected more by sudden changes in elevation data. It is even more critical to know where the vehicle can be driven without tipping over or encountering an impassable obstacle.
Simulating navigation in crew-based vehicles
Historically, the simulation of ground vehicles has separated the physical modeling of the host vehicle and its environment from the visual representation portrayed in its corresponding operator training system. This separation was needed because typical visual systems could not cope with the amount of information required in both the physical modeling of the vehicle and terrain environment. The criterion defining this separation is the relationship between the wheelbase of the vehicle being simulated and the resolution of the underlying terrain. Often the simulation host model running the vehicle physics model ran its own higher resolution version for the terrain with a lower resolution version on the visual systems. In such situations, anomalies occurred for the driver; subtle driving cues that were experienced with high-resolution elevation data were not visible in the associated visual system if the two systems use different databases. Navigating through densely featured urban enviroments is one example of where having an accurate terrain representation is especially important.
Another example of the effect of elevation resolution on suspension
As you cycle through the images of the 8-wheeled Stryker vehicle below (by clicking them) you can see how the suspension system of the vehicle reacts more dramatically to the higher resolution elevation data used to create the underlying terrain.
MetaVR model of Stryker vehicle. Click the image to cycle through the effect of the elevation data on suspension.
The higher resolution 1-meter terrain provides more accurate dynamics feedback which in turn gives a more accurate prediction of how the real vehicle will behave during actual use. This provides a better evaluation model for simulation-based acquisition and a more accurate model for training the vehicle operators.
For a realistic simulation experience, the terrain model must represent a surface at a finer resolution than the vehicle's wheelbase, what we refer to as "better-than-wheelbase terrain". Currently an important transition is occurring in modeling and simulation of ground vehicles. PC-based image generators that have extremely wide data bandwidths are reducing the separation between the physical model and the visual model, which in turn greatly improves the accuracy of the simulated behavior. MetaVR's real-time image generator, VRSG™, is capable of visualizing the terrain geometry at resolutions commensurate with the physical models.
Example of high resolution terrain elevation under the vehicle
As you cycle through the images below you can see that the terrain support points at 1-meter resolution lie within the confines of the wheelbase of the vehicle and therefore provide very high-fidelity terrain interaction for the vehicle suspension system used to create the underlying terrain.
Wireframe view underneath the Stryker vehicle. Click the image to cycle through the increasing higher-resolution elevation data.