Agents with Adjustable Autonomy
Papers from the AAAI Spring Symposium
David Musliner and Barney Pell,Cochairs
Adjustable autonomy means dynamically adjusting the level of autonomy of an agent depending on the situation. For real-world teaming between humans and autonomous agents, the desired or optimal level of control may vary over time. Hence, effective autonomous agents will support adjustable autonomy. This contrasts with most work in autonomous systems, in which the style of interaction between the human and agent are fixed by design. The adjustable autonomy concept includes the ability for humans to adjust the autonomy of agents, for agents to adjust their own autonomy, and for a group of agents to adjust the autonomy relationships within the group. Effective adjustable autonomy minimizes the necessity for human interaction, but maximizes the capability for humans to interact at whatever level of control is most appropriate for any situation at any time. This symposium will bring together academic and industrial researchers from diverse fields such as classical and reactive planning, mixed-initiative systems, distributed AI, robotics, intelligent user interfaces, and multi-agent systems. Specific topics in the technical report include human-involved procedure execution; mixed-initiative planning; monitoring and situation awareness; anytime behavior modification of agents; maximizing effective collaboration between humans and agents; human-centered autonomous systems; understanding or controlling agents having adjustable autonomy; validation and verification of agents having adjustable autonomy; resource and constraint-based autonomy; levels of autonomy; safety and fault-tolerance; mission criticality; long-distance interaction; modeling and predicting behavior; reasoning about communication; prioritized tasking; case studies on problems in mixed human-machine operations; associate systems and overlapping competence; command-by-exception; and flexible team organization between humans and agents.