Curating User-Defined Interface Maps for Robot Teleoperation

Area of Science:

• Robotics
• Human-Computer Interaction
• Assistive Technology

Background:

• Traditional assistive robot interfaces use fixed command maps, limiting personalization.
• A one-size-fits-all approach does not cater to individual user preferences or capabilities.

Purpose of the Study:

• To develop novel methods for eliciting user-defined control maps for assistive robots.
• To address data quality issues inherent in user-centered control map design.
• To evaluate the efficacy of proposed methods in a user study.

Main Methods:

• Elicitation of user-defined control maps through a user study.
• Identification and mitigation of control signal data issues.
• Development of signal filtering and synthetic data generation techniques.

Main Results:

• Experimental evaluation using a powered wheelchair and robotic arm across four interfaces.
• Demonstration of differing suitability of user-defined maps for various interface-platform combinations.
• Analysis of errors in dataset creation and the impact of postprocessing.

Conclusions:

• User-defined, data-driven control maps offer a personalized alternative to fixed interfaces.
• Proposed methods effectively address data quality challenges in user-centered assistive robot control.
• Findings provide insights into optimizing user-defined control maps for diverse robotic platforms.