Cause and effect: the case for better correlated training environments

For truly immersive collective training, the concept is simple: you want a fair fight for everyone involved. This is especially true in regards to simulated training. When using different virtual environments, they need to be correlated otherwise discrepancies arise in outcomes that undermine the value of the training.

A tree blocking the view of a user in one environment and not in another, for example, will cause problems and a break down in communication. Should these same users get ambushed during a virtual training exercise, they’ll need correlated lines of sight and head direction so both know which direction to look in. Their weapon positions should sync so both are aiming accurately at the same target – which, again, should look the same across different environments.

VBS4 provides immensely rich and detailed terrains, coupled with dynamic entities, which makes solving these correlation challenges harder when using different virtual environments together. But the correlation challenge needs solving, given that VBS4 has been selected as the Defence Virtual Simulation (DVS2) solution for UK. Its widespread use and usefulness means we must explore the art of the possible to provide fully correlated environments in which both a foot soldier and a tank crewman who may be in a different virtual environment can train fairly and to the best of their ability.

To do this, we’re going beyond network standards to capture the level of nuance required for better, fairer and more effective training. Skyral will integrate with VBS4 at an API level, using the VBS4 SDK, which comprises BISim’s extensible API library, as part of its toolbox.

The end goal is to harness VBS4 as a window onto a wider correlated virtual environment, composed of specialist technologies and applications that enable humans-in-the-loop to experience entirely new aspects of training:

• Accurate visual representation of modelling and simulation involving civilian sentiment and infrastructure effects;

• Massive scale and density comprising upwards of 1m complex entities;

• The exchange and tracking of nuanced data for in-scenario monitoring control and analysis to better understand training performance both as it happened and as it could be conducted in future.

Simulating the information domain: synthetic sentiment, accurate animations and infrastructure modelling

The real world is complex. So are the humans that inhabit it. Collective training – whether live, virtual or constructive – needs to reflect this complexity to give trainees as realistic an experience as possible.

So we turn to modelling and simulation. Non-player characters (NPCs) in VBS4 will exhibit behavioural visual traits based on sentiment that’s been generated by the complex interaction within applications such as Montvieux’s ASAIR and Improbable’s civilian population models, coupled with critical national infrastructure (CNI) networks and information spread effects through those networks. Should an exercise facilitator inject a CNI event, VBS4 can make use of Skyral’s infrastructure network modelling capabilities.

Once triggered, the aforementioned foot soldier and tank crewman will witness the visual manifestation of sentiment change in the affected civilian population controlled by Skyral, but visualised in VBS4. Alongside trainees at HQ level who can see wider changes across the region, they must decide on the best course of action according to their skillsets and experience. And that’s the beauty of a correlated operating environment composed from multiple specialist technologies in which cause and effect impacts training as part of a larger feedback loop.

A (no so) small world: maximum scale and density

Complex NPCs are essential for truly immersive collective training. In the case of VBS4’s integration with Skyral, there will be a lot of them: over 1m, to be exact, each optimised for unprecedented scale and running in the cloud. They’ll exhibit a range of specific traits, such as social media use, connection to telecom networks, fear for life and sentiment towards local government. These will be updated at different rates by multiple models, simultaneously across the whole population.

Improbable’s games heritage gave us a head start in achieving this. Advances made by our Games Technology Unit in the development of ScavLab and Project Morpheus have extended the technological limits of scale and density, with minimal latency and network burden. These same technological advances also underpin Skyral. For the first time, users can experience a window onto a vast, densely populated world visualised in VBS4, tailored to their exact objectives and at the scale needed to accurately reflect real-world operations.

Harnessing data for in-scenario monitoring control and analysis

Deeper levels of integration between specialist technologies means we can tap into more nuanced data for richer, more immersive training.

To return to my opening example: standalone VBS4 can simulate the look, feel and formation of a military convoy. By integrating VBS4 and Skyral, we can go a step further by making use of better data to set up triggers that alert users to very granular in-scenario events.

Let’s say this convoy forms an essential part of a resupply to front line troops. Naturally, VBS4 has a trigger to alert trainees if a vehicle’s fuel level has become critically low. This feature is fed directly into Skyral’s canonical view of the operating environment.

The prospect of the vehicle’s paralysis means trainees will have to collectively stop, take stock and figure out the next best course of action. Thanks to our work with 4C Strategies, events such as fuel states, civilian casualties or sentiment changes can be automatically flagged in Exonaut so observer-mentors know to analyse trainee response during after-action review.

Remember: the Russian operation stalled on Kyiv Oblast precisely because of this logistical oversight. By simulating situations like it, trainees are better prepared and more confident in the face of similar logistical challenges on real-world operations.