Human-automation collaboration in occluded trajectory smoothing

Abstract

Deciding if and what objects should be engaged in a Ballistic Missile Defense System (BMDS) scenario involves a number of complex issues. The system is large and the timelines may be on the order of a few minutes, which drives designers to highly automate these systems. On the other hand, the critical nature of BMD engagement decisions suggests exploring a human-in-the-loop (HIL) approach to allow for judgment and knowledge-based decisions, which provide for potential automated system override decisions. This BMDS problem is reflective of the role allocation conundrum faced in many supervisory control systems, which is how to determine which functions should be mutually exclusive and which should be collaborative. Clearly there are some tasks that are too computationally intensive for human assistance, while other tasks may be completed without automation. Between the extremes are a number of cases in which degrees of collaboration between the human and computer are possible. This thesis motivates and outlines two experiments that quantitatively investigate human/automation tradeoffs in the specific domain of tracking problems. Human participants in both experiments were tested in their ability to smooth trajectories in different scenarios. In the first experiment, they clearly demonstrated an ability to assist the algorithm in more difficult, shorter timeline scenarios. The second experiment combined the strengths of both human and automation to create a human-augmented system. Comparison of the augmented system to the algorithm showed that adjusting the criterion for having human participation could significantly alter the solution. The appropriate criterion would be specific to each application of this augmented system. Future work should be focused on further examination of appropriate criteria.