A great many computer-based training systems are being built and experiments in new training techniques are being conducted, yet many critical questions remain unaddressed and appear resistant to current approaches. After tackling the research questions posed during DARPA's DARWARS program and spending years building training systems and developing techniques to meet particular training requirements, the authors felt that there were fundamental, cross-cutting challenges that needed to be met in order to make further progress toward delivering comprehensive training solutions. Our hope is that these challenges, if explicitly formulated and directly addressed, will provide a vision for the future of computer supported training. This short manifesto sets forth ten challenges, each of which could serve as the central focus of an R&D program:
1. Training Untrainable Skills - Training the intangibles: leadership, adaptability, resilience, or vigilance.
2. Practical, Collective Training - Training for large groups distributed in time and space.
3. Training to Learn - "Meta" training on the skills of acquiring expertise.
4. Keeping Training Current - Continually updating content.
5. Capturing and Transferring Experience - From the few to the many.
6. Training Each Other - Crowd-sourcing training; enabling peer training.
7. Training to Remember - Solving the skill retention problem.
8. Ubiquitous Training - Making training a natural part of task performance.
9. Persistent Mentoring - Providing career-long guidance.
10. Training to Excel - Pushing learners toward their top performance; aiming at excellence rather than training to standard.
In spite of some progress in addressing these items in the course of many different efforts, we contend that programs explicitly focused on these challenges would provide insight, progress and capability that will not be developed if we only continue with research on specific techniques, or the development of individual training systems.