Modern warfare stresses the importance of maintaining the required proficiency level for todays and future generation air force pilots. These pilots must be able to execute various military missions with rapid deployments, which demands ongoing training. However, today’s training is mostly standardized per community and every pilot receives roughly the same training, despite having different learning curves and training needs. As a result, the total training is inefficient, both from a corporate and individual perspective, and does not achieve the best possible operational readiness.

Recent advances in technology, such as artificial intelligence, data science, learning analytics, simulation, x-realities and the internet-of-things, have enabled the implementation of new training strategies, such as performance-based, personalized and adaptive training. These strategies aim to utilize available live and virtual training assets optimally, in order to main the highest possible pilot performance. The digital era allows us to create a service-oriented training ecosystem, which objectively tracks, analyzes and utilizes training data to offer each individual pilot, on-demand, the training most beneficial to their learning curve.

The concept of data-driven training combined with a service-oriented training provision approach is referred to as Training as a Service (TaaS). In this paper, we elaborate on this concept from a technical perspective, including the required architecture for enabling TaaS. The infrastructure consists of several elements: a training element that controls all training-related parts, including the Mission Environment as a Service (MEaaS) and adaptive scenarios; an analysis suite that performs logging and data analysis; a dashboard to control, manipulate and visualize training; and a training orchestrator as the backbone of the infrastructure. We illustrate this with a concept demonstrator developed by Royal NLR, which includes services for training orchestration and scheduling, digital training syllabi, xAPI2 learning lockers, dashboards, learning analytics, cognitive load measurements, and scenario configuration with adaptive Computer Generated Forces (CGF).

Keywords
ARCHITECTURE, LEARNING ANALYTICS, M&S AS A SERVICE, PERSONALIZED TRAINING, SERVICES-BASED SIMULATION

Additional Keywords