Automated patient-robot assignment for a robotic rehabilitation gym: a simplified simulation model

Area of Science:

• Robotics
• Rehabilitation Engineering
• Artificial Intelligence

Background:

• Robotic rehabilitation gyms offer positive outcomes and cost-effectiveness by allowing therapists to supervise multiple patients.
• Automated systems are needed to dynamically assign patients to robots for optimized rehabilitation.

Purpose of the Study:

• To develop and evaluate a mathematical model for dynamic patient-robot assignment in robotic rehabilitation gyms.
• To maximize total skill gain across all patients and skills within a session.

Main Methods:

• A simplified mathematical model of a robotic rehabilitation gym was created.
• Mixed-integer nonlinear programming algorithms were used to find optimal assignment and training solutions.
• Scenarios varied patient/robot numbers, training durations, and complexity levels.

Main Results:

• Optimization models significantly outperformed baseline schedules (staying on one robot or switching halfway).
• The disjunctive model yielded higher skill gain than the time-indexed model.
• Optimization duration increased with more patients, robots, and time steps.

Conclusions:

• Intelligent patient-robot reassignment enhances skill acquisition in multi-patient, multi-robot settings.
• This decision support system can improve the efficiency of technologically aided rehabilitation.
• While scenarios were simplified, the findings support the potential of automated assignment in future robotic rehabilitation gyms.