The AMIRO Social Robotics Framework: Deployment and Evaluation on the Pepper Robot.

Area of Science:

• Robotics
• Human-Computer Interaction
• Artificial Intelligence

Background:

• Social robotics offers economic benefits in retail, entertainment, and active and assisted living (AAL).
• There is a growing demand for affordable, robust social robots providing specific assistance.
• Existing frameworks may lack modularity or specific focus for assistive care scenarios.

Purpose of the Study:

• To introduce the AMIRO social robotics framework, engineered for modularity and robustness in assistive care.
• To detail the framework's components, including navigation, person detection, natural language interaction, and activity recognition.
• To demonstrate a platform-independent implementation using the Robot Operating System (ROS).

Main Methods:

• Development of a modular framework with distinct robotic service modules.
• Implementation of a platform-independent architecture utilizing ROS.
• Quantitative evaluation of individual module performance across various settings.
• Deployment on the Pepper social robot platform to showcase integrated functionality.

Main Results:

• The AMIRO framework provides essential robotic services for assistive care.
• Quantitative evaluations demonstrate the performance of each module.
• Successful deployment on the Pepper robot highlights the framework's practical application.
• Development of a complex user interaction scenario utilizing all framework modules.

Conclusions:

• The AMIRO framework is a versatile and robust solution for social robotics in assistive care.
• The modular design facilitates adaptation and improvement of individual components.
• The platform-independent ROS implementation enhances its applicability across different robotic systems.
• The framework enables the creation of sophisticated human-robot interaction scenarios for enhanced user assistance.