About 60% of children with cerebral palsy (CP) have impaired arm function. Improving arm function requires hundreds of repetitions per day, which is impossible in a single clinical visit. Thus, therapeutic exercises should be sustained in the home environment; however, the compliance in performing home exercise is low due to poor motivation, boredom, and forgetfulness. A better home program is needed. The objective of this project is to examine the effect of our developed "THRIVE" system (Therapeutic Humanoid Robot In Virtual Environment: the combination of robot with virtual reality games), which can provide a motivating and tailored upper-extremity intervention program with instant feedback, to improve arm function in children with CP at their home. The investigators believe the newly developed "THRIVE" system can increase children's engagement and motivation in home exercises as the robot is their playmate to demonstrate and correct their movements. The investigators will also interview parents and children with CP to understand their impression of using technology at home to shape the intervention. The expected outcome is that children with CP receiving "THRIVE" will improve their arm function more and have better engagement than those who receive VR alone immediately after intervention and at follow-up. The long-term goal is to have the "THRIVE" system be the optimal home exercise platform as it can provide challenging but motivating exercises to improve children's arm function while assisting parents in supervising their children with CP to complete home exercises.

Inclusion Criteria:

1. Participants are between the ages of 5 and 21 years;
2. diagnosed with spastic CP;
3. have a manual ability classification system (MACs) level I-III;
4. able to sit with trunk supported;
5. are able to reach forward for more than half of their arm length;
6. are able to comprehend and complete a three-steps commands;
7. are able to see a TV screen (with or without corrected vision); and
8. their primary caregiver is willing to follow the desired intervention dosing and all evaluation measurements.

Exclusion Criteria:

1. they have received surgery or botulinum toxin type A injection in the training arm within the preceding 4 months or are scheduled to receive it during the planned intervention period, or
2. if they have a severe attention deficit or uncontrolled epilepsy, which may possibly be triggered by the light or sound of the video games.

Real user therapy assessment and outcome data and intervention log will be possibility shared to a NIH-funded open video-based data-sharing system for developmental science (i.e., Databrary). Databrary is one of the first large-scale repository for sharing video data and related information so the researchers can share and view on another's databases to promote greater transparency and peer oversight into data collection methods and measurement. All the videos collected through Kinect camera during interventions and evaluations will be de-identified first before sharing to Databrary. A new written permission will be obtained from the parents of children with CP before initiating the data sharing process. PI, Co-PI, and research team members will contact the parents to explain the purpose of data sharing. If parents show their initial agreement, a de-identified video will send it to the parents for review before uploading and sharing to Databrary.

9/1/2027 - after the data collection is completed and data have been analyzed and disseminated. Data will be available until 12/31/2035.

Reserachers who also sign up to participate in datasharing in Databrary.