Evaluation of VR-Based Speed and Direction Control Strategies for Pediatric Robotic Gait Training

Background Cerebral Palsy (CP) is a neurological disorder affecting motor control, often requiring lifelong rehabilitation. Robot-Assisted Gait Training (RAGT) shows promise for pediatric rehabilitation, but maintaining engagement remains challenging. Virtual Reality (VR) offers potential for creating motivating therapeutic environments; however, successful integration with RAGT systems requires careful technical validation before clinical deployment. This study presents the development and technical validation of a VR-integrated RAGT platform that allows children to actively control their walking speed and direction, assessing system safety and user impact. Methods We developed an integrated system combining the Discover2Walk cable-driven robotic platform with VR environments. The system implements three control modalities: (i) self-paced speed control using pelvic position and interaction forces, (ii) head-orientation-based directional control via VR headset and (iii) dual control combined speed and direction. Technical validation was conducted with fourteen Typically Developing Children (TDC) who tested three progressively VR games designed to evaluate each strategy independently and combined. System performance was assessed through objective metrics including Speed Tracking Error (STE) and Kinematic Tracking Error (KTE). A case study with one Cerebral Palsy Participant (CPP) provided preliminary insights into system usability with the target population. Safety was monitored using the Pediatric Simulator Sickness Questionnaire, and user experience was evaluated through enjoyment ratings. Results Technical validation demonstrated robust system performance. The TDC group successfully tracked target speeds with mean errors of 10% for faster and 20% for slower conditions. Navigation tasks showed comparable accuracy between fixed-speed (KTE: 0.73 m) and dual-control (0.71 m) conditions. All participants completed the goal-oriented tasks with high success rates. The system demonstrated excellent safety profiles with minimal cybersickness symptoms (all scores < 1 on a 0–6 scale) and high user acceptance (enjoyment ratings > 9/10). The case study revealed the system's accessibility for children with CP, though with expected performance variations compared to TDC. Conclusion The developed platform enables intuitive, engaging, and safe interaction through validated self-paced speed control and head-orientation navigation methods. High user acceptance and minimal adverse effects support the system's readiness for expanded clinical validation studies. These technical foundations establish a framework for future research investigating therapeutic efficacy in larger clinical populations.