Ernest V. Cross II, Jimena Guallar-Blasco, Matthew Miller, Leonard Eusebi

Interservice/Industry Training, Simulation, and Education Conference (I/ITSEC), Orlando, Florida (3 December 2025) 

In many contexts, trainees need individualized, on-demand content to refresh skills. To provide this capability, we
designed an intelligent user interface (IUI) that uses a context -aware system for just-in-time training to convert dense,
static technical manuals into dynamic, actionable training content. Traditional technical manuals and training materials
are often lengthy and difficult to navigate, making it hard for users to quickly access the critical information they need,
whether for routine maintenance or unexpected repairs. Furthermore, existing training materials have been bound by
fixed formats and device dependencies, which limits accessibility and flexibility.

This research presents preliminary work on a system designed to overcome the usability constraints of current
technical manuals, which impede efficient information retrieval during both standard operations and time -sensitive
scenarios. Our system employs computational methods including deep learning, natural language processing
techniques, and multimodal artificial intelligence to extract critical features of technical content—including procedural
workflows, safety protocols, equipment specifications, and multimedia resources—from heterogeneous
documentation repositories. This enables the creation of customized training modules through an integrated
multimodel architecture that evaluates task requirements, training content, user expertise, environmental constraints,
and device capabilities. This context-aware framework enables the system to deliver personalized training
interventions—comprehensive guides for novices, concise refreshers for experienced personnel, and just -in-time
support for immediate operational support—optimized for the specific operational setting and context (e.g., noise
level, available interaction modalities, time constraints, and other factors).

We will describe our initial demonstration of a capability to systematically structure and remix training content from
diverse sources into a standardized, machine-interpretable format suitable for dynamic distribution across multiple
technological platforms, including mobile devices and tablets, desktop interfaces, and immersive reality environments.
We will also discuss future applications of this work in both defense operations and commercial sectors, such as
automotive repair, maintenance services, and safety procedures.

For More Information

To learn more or request a copy of the paper, contact Vince Cross.

(Please include your name, address, organization, and the paper reference. Requests without this information will not be honored.)