Search for: All records

Intelligent multi-purpose robotic assistants have the potential to assist nurses with a variety of non-critical tasks, such as object fetching, disinfecting areas, or supporting patient care. This paper focuses on enabling a multi-purpose robot to guide patients while walking. The proposed robotic framework aims at enabling a robot to learn how to navigate a crowded hospital environment while maintaining contact with the patient. Two deep reinforcement learning models are developed; the first model considers only dynamic obstacles (e.g., humans), while the second model considers static and dynamic obstacles in the environment. The models output the robot’s velocity based on the following inputs; the patient’s gait velocity, which is computed based on a leg detection method, spatial and temporal information from the environment, the humans in the scene, and the robot. The proposed models demonstrate promising results. Finally, the model that considers both static and dynamic obstacles is successfully deployed in the Gazebo simulation environment. 

The paper proposes a robot learning framework that empowers a robot to automatically generate a sequence of actions from unstructured spoken language. The robot learning framework was able to distinguish between instructions and unrelated conversations. Data were collected from 25 participants, who were asked to instruct the robot to perform a collaborative cooking task while being interrupted and distracted. The system was able to identify the sequence of instructed actions for a cooking task with the accuracy of 92.85 ± 3.87%. 

Robots have the potential to assist people in daily tasks, such as cooking a meal. Communicating with the robots verbally and in an unstructured way is important, as spoken language is the main form of communication for humans. This paper proposes a novel framework that automatically generates robot actions from unstructured speech. The proposed framework was evaluated by collecting data from 15 participants preparing their meals while seating on a chair in a randomly disrupted environment. The system can identify and respond to a task sequence while the user may be engaged in unrelated conversations, even if the user’s speech might be unstructured and grammatically incorrect. The accuracy of the proposed system is 98.6%, which is a very promising finding.