Automation systems that can find collision-free paths in uncertain environments with dynamic obstacles have highly significant applications, including aerial navigation, satellite coordination, and self-driving cars. However, many of these systems require human monitoring for special cases the automation can not handle. For these cases, an optimal experience results from effective collaboration between the human and automation. Our research focuses on finding ways to adapt technologies for successful human-automation collaboration.
For this purpose, we have developed Busy Beeway. Busy Beeway is a mobile game in which the player must guide the bee avatar, Beelinda, to a goal while avoiding stochastic obstacles (usually wasps). Depending on the mode, the player may complete a level solely or with varying levels of automation guidance. Automation methods are derived from our Moving Obstacle Avoidance Project.
Through user studies of Busy Beeway, we seek to answer questions such as:
How much is the human willing to rely on the automation?
What game states cause the human to take control?
What are effective mechanisms for conveying the guidance (including uncertainty) information to the player?