Shaping of Shared Autonomous Solutions With Minimal Interaction.

Area of Science:

• Robotics
• Artificial Intelligence
• Human-Computer Interaction

Background:

• Specifying problems for autonomous systems is a key challenge in shared autonomy.
• Existing methods often require extensive human input or complete information transfer.

Purpose of the Study:

• To introduce and evaluate the Interactive Shared Solution Shaping (IS3) paradigm.
• To demonstrate that minimal human interaction can optimize autonomous planning.

Main Methods:

• Developed the IS3 framework for human-robot collaborative planning.
• Applied IS3 to resource-constrained mobile search and surveillance tasks.
• Conducted experiments to assess route generation and performance.

Main Results:

• IS3-generated routes improved target detection performance and reduced variance.
• Optimized the trade-off between performance and interaction cost.
• Demonstrated real-world feasibility with a robot operating within budget constraints.

Conclusions:

• IS3 enables efficient human guidance of autonomous systems.
• Minimal interaction effectively shapes autonomous planning for complex tasks.
• The IS3 paradigm offers a practical approach to shared autonomy in mobile robotics.