Abstract
The scheduling of pilot training events is a highly complex and data-intensive process that relies heavily on manual heuristics, often resulting in suboptimal outcomes. Recent advancements in predictive algorithms have demonstrated the ability to forecast student proficiency based on training exposure. Combined with data on current student performance and projected resource availability, the foundational elements exist to enable a dynamic scheduling approach tailored to both individual and enterprise-wide needs.
At the core of this approach is a Skill Attainment model that accounts for the complexity of skills, prior exposure, and the quality and quantity of proposed training events. This model provides a directional estimate of how proficiency will improve with training or degrade due to inactivity. Applications of a Skill Attainment model in Air Force training command and Naval strike contexts will be discussed.
Complementing this is the Training Enterprise Model (TEM), which represents the capacity of the broader system to support training schedules—incorporating hard assets such as classrooms, aircraft, and runways, alongside human resources like instructors and maintainers. Examples of the TEM from both Air Force and Navy pilot training schools is used for calibration and verification.
Together, these models offer a holistic view of the training environment. However, the decision space is massively multi-dimensional, and the cadence of real-world training demands automation to avoid overwhelming human schedulers. A non-convex optimization framework is applied to generate automated schedule suggestions that maximize student cohort proficiency without exceeding enterprise capacity or causing unintended ripple effects across the system.
This integrated approach has been applied across both active-duty training pipelines and introductory student syllabi, yielding near-term optimal schedules that reduce instructor workload and improve force readiness. By combining predictive modeling with automated optimization, this method represents a transformative shift in how military training enterprises can balance individual skill development with operational constraints.