Book
Robot-Assisted Rehabilitation of Children with Cerebral Palsy
Authors: Prashant Kumar Jamwal
Publishing Date: 06-11-2026
ISBN: 978-81-975670-9-4
Preface
Cerebral palsy is the most common cause of physical disability in childhood. Its consequences reach well beyond movement: many children also live with impairments of balance, speech, communication and cognition, and these often matter as much to families as walking does. Rehabilitation for these children is long and labor-intensive, it is unevenly available, and the evidence base for many interventions is still thin. Robotics and artificial intelligence promise to make therapy more intensive, more measurable and more consistent. They also risk adding expensive equipment that has not been tested with the children it is meant to help. I have worked in rehabilitation robotics since 2007, and over those years I have come to think that the second risk deserves as much attention as the first promise. This book reports what one national program learned while trying to keep that promise without overstating it. Program BR27199433, funded by the Ministry of Science and Higher Education of the Republic of Kazakhstan and co-financed by Zein Invest LLP, brought together engineers from the School of Engineering and the Center of Excellence in Medical Robotics and Research at Nazarbayev University, clinicians from the University's School of Medicine and from the National Center for Children's Rehabilitation (NCCR) in Astana, and special-education researchers from the Graduate School of Education. NCCR receives roughly 300 children a month, and its willingness to open its therapy rooms to research made it possible to work with children at a scale that is rare in paediatric rehabilitation robotics: 170 children in the social-robot studies, 25 in the perturbation-training trial and 25 in the gait-robot trial. The program developed or adapted four devices. The Gait Rehabilitation Robot is a hip–knee exoskeleton that walks with the child on a treadmill and measures, through a load cell at each joint, how much of the effort is the child's own. The adaptive body-weight support system, designed by our team, uses a compliant pneumatic drive to unload the child by a prescribed and individually chosen amount. The Robotic Perturbation Trainer is a cable-driven robot that pulls briefly and unpredictably at the pelvis during walking, so that children can practise recovering their balance in a safe setting; its final version can perturb in all three planes. The fourth device is QTrobot, a commercial social robot that we taught to speak Kazakh as well as Russian and for which we wrote bilingual therapy scripts for communication, cognition and social skills. Alongside the devices, the program built two software platforms: GARAS, which turns the gait robot's sensor data into a Gait Recovery Index and rule-based therapy recommendations, and the NCCR Rehabilitation Platform, which brings clinical and sensor data together for documentation and research. The book is organised in four parts. Part I sets out the clinical problem, the international state of the art and the methodology of the program. Part II describes each device in turn: its rationale, design, control and safety architecture, and the protocols by which it is used. Part III describes the software through which the devices produce data and indices for clinicians. Part IV presents the clinical evidence gathered so far, discusses what it does and does not show, and considers the route from laboratory prototype to registered medical device and routine service. Throughout, I have tried to keep apart three kinds of material that are easily confused in a field driven by technology: what has been published after peer review, what is still under review or in preparation, and what is still pending. Each chapter states the status of the results it reports. Several studies described here are pilot or feasibility studies with small samples; I present them as such, and I show, with sample-size calculations based on our own pilots, how much larger a confirmatory trial would need to be. Where an analysis method is demonstrated before trial data are available, the demonstration is labelled as such at the point of use. One point needs to be stated plainly. At the time of writing, the randomised trial of the Gait Rehabilitation Robot, registered as ACTRN12626001203347, and the perturbation-training trial are still in progress. Enrolment for the gait trial began in September 2026, and robot-assisted sessions began in October 2026 after the final ethics conditions for the device had been met. The results of these two trials are therefore pending, and this book describes their designs and analysis plans rather than their outcomes. I would rather publish an honest account of a program in progress than a premature claim of benefit. The book is written for researchers and graduate students in rehabilitation engineering, for clinicians and therapists who work with children with neurological disabilities, and for those who plan and fund rehabilitation services, particularly in Central Asia. I hope it does justice to the children and families who gave their time to this work.
Cite as
Kumar Jamwal, Prashant. 2026. Robot-Assisted Rehabilitation of Children with Cerebral Palsy: Devices, Data and Clinical Evidence from Kazakhstan (Computing and Intelligent Systems). Soft Computing Research Society.