Transcending Reality to Protect Our Armed Forces

As a society, we rely on data for everything from finding the best route to the grocery store to determining if we have reached the right level of fitness for the day. Because we are so easily exposed to these data sources, there’s a misconception that incorporating data science into military training is as simple as buying a headset and playing a game. But, as an organization, it can be overwhelming to think about all the infrastructure that needs to be in place to collect and use this data in an impactful way.

If built strategically, organizations can create a foundation — literally and figuratively — as they change and continue to mobilize. That technological foundation needs to be an open and non-proprietary framework for us to adopt and adapt, while addressing challenges such as security and privacy. If we direct our energy the right way, it will give us a solution that isn’t as cumbersome as the one we currently rely on, delivering training simulations right to the fingertips or headsets of our service members without the months-long or year-long delay for new training simulations that exists today.

Our idea of the perfect military readiness platform allows new data to be ingested, whether from training simulations, biometrics, or other data sources. The goal, of course, is to gain insights to improve the readiness of our Armed Forces and reduce the risk of loss of life or injury in the future. We can move from having to manipulate data to work with simulations to having simulations that are able to ingest data from any data source. In effect, we’re transforming simulation-ready data to data-ready simulations.

With Immersive technology, we are collecting a ton of data — everything from the user’s physical behavior and what they are looking at to how fast they are reacting. To ignore this data would be a huge missed opportunity, and one that we believe would hinder the advancement of immersive technology and its ability to truly change the way we live in and interact with our environments.

This is where machine intelligence — including computer visioning, neural networks, quantum computing, deep learning, and artificial intelligence – comes in. We need to start building machine intelligence models that enhance military training by mimicking human behavior learned from service member interactions within the simulation.

Machine learning can create “smart agents” that respond to the trainee’s voice, actions, and commands and execute the proper course of actions based on the virtual training scenario. In the current live training environment, these characters need to be either played by trainers at a high cost or neglected completely. Smart agents ingest data from the trainee’s interaction and provide the feedback at a lower cost.

Machine learning increases the realism of scenarios by either augmenting or replacing the scripted step-by-step scenarios with more realistically generated randomness – just like warfighters’ experience in the field. Utilizing these technological advances, training can more closely match the experience of real world situations while still holding on to core lessons.

In addition to creating more realistic interactions, machine intelligence allows us to analyze the vast amount of data collected to provide real-time feedback to the users or the organizations training them. This is a critical part of enhancing decision making, training, and optimization of spaces. As technology advances to include human-brain and computer interfaces, allowing us to control devices with our minds or to implant technologies into our bodies to more efficiently operate, machine intelligence will evolve as well to further power these immersive technologies.

Without a comprehensive approach to immersive training — one that converges the digital and physical worlds by layering machine intelligence and data science on to training simulations and operations — our soldiers will continue to “game the game” instead of gaining the realistic training they need.