Learning Dynamic Patient-Robot Task Assignment and Scheduling for A Robotic Rehabilitation Gym

Area of Science:

• Robotics
• Artificial Intelligence
• Rehabilitation Medicine

Background:

• Robotic rehabilitation gyms enable group therapy with multiple patients and robots.
• Optimizing patient-robot assignments can enhance the effectiveness of group rehabilitation.

Purpose of the Study:

• To develop an automated system for dynamic patient-robot assignment in robotic rehabilitation gyms.
• To improve group rehabilitation outcomes through intelligent assignment strategies.

Main Methods:

• A neural network classifier was developed to predict patient-robot assignment priority based on task success rates.
• A robotic gym simulator was created to generate a synthetic dataset for training the neural network.
• The system dynamically assigns patients to robots to optimize the overall group rehabilitation outcome.

Main Results:

• The developed automated system effectively assigns patients to robots.
• The simulation results indicate that the system achieves patient training outcomes comparable to those determined by a human domain expert.
• The neural network classifier, trained on synthetic data, proved capable of making effective assignment decisions.

Conclusions:

• Dynamic patient-robot assignment using a neural network classifier is a viable strategy for optimizing robotic rehabilitation.
• The automated system shows promise for enhancing the efficiency and effectiveness of group robotic rehabilitation.
• Further development and validation in real-world settings are warranted to fully realize the potential of this approach.