AI-DRIVEN TRAINING SCHEDULE AUTOMATION AND OPTIMIZATION

Pilot training, pilot production, and mission readiness requires the management and orchestration of resources, people, and training curricula. Continuous pressure to increase pilot production capacity and maintain or improve mission readiness has led the USAF to embrace new training technologies and experiment with alternate training curricula that are able to meet these demanding requirements. Continued modernization and structuring of the force increases capabilities but the overlooked effective and efficient training management can be a real force, and capability, multiplier, can improve training effectiveness, reduce costs, deliver more and better-trained pilots, and maintain a ready force. Legacy systems are old, not cloud-based, not updatable in real time, and limit the degree of automation and integration of new technologies. We propose the adaptation of the civilian Flight Optimal platform to the needs of the USAF. Flight Optimal is a full-service training management system built on an automation and optimization platform driven by a combination of optimization algorithms, artificial intelligence-based projection and forecasting system, and smart automation architecture to generate and update flight schedules based on individualized training goals, multiple syllabi, training progression, real-time availability and status of resources, students, and instructors, and user-defined and automated prioritization criteria of training requirements. The proposed effort directly supports the USAF goals for digital transformation and operational efficiency through the delivery of advanced computing and software. The automation and optimization of the scheduling process delivered by Flight Optimal provides a cloud-based system for the storage, data processing, and advanced computing of scheduling solutions. The modular architecture enables user communities to utilize operation-specific algorithms and capabilities without additional development and continuously receive updates on new computing capabilities at the enterprise level. The API architecture allows for seamless integration with existing databases to reduce duplication and leverage the strengths of Flight Optimal.  The automation and optimization of the scheduling process and resource management directly increases training capacity by addressing critical training bottlenecks, by increasing the utilization rates of available resources, and ultimately increasing mission readiness. The civilian platform has already demonstrated multi-fold reduction in workload, up to 25% increase in the training pipeline, and above 85% scheduling effectiveness.