The U.S. Army Research Laboratory and computer science researchers at North Carolina (NC) State University entered an agreement to develop a program called DeepGen to generate customizable virtual training scenarios.

The scenarios created by DeepGen will run on the Bohemia Interactive Simulations-developed Virtual Battlespace 3 (VBS3), a program already used for infantry training by the U.S. Army. The team is also working with Intelligent Automation Inc. to address variables such as weather, visibility, physical environment and more in training scenarios.

“We want DeepGen to create scenarios that not only incorporate targeted learning objectives, but reflect the strengths and weaknesses of individual trainees, based on each trainee’s demonstrated competencies up to that point,” Said Bradford Mott, an NC State research scientist involved in the project.

To get started, the DeepGen researchers will focus on training Army personnel on how to call in artillery fire.

“The DeepGen system will also continue to grow and change over time, using a machine learning technique called deep reinforcement learning to improve the scenarios it creates based on how trainees interact with each scenario,” said Jonathan Rowe, a research scientist on the NC State team.