PROJECT SUMMARY Stroke is the leading cause of serious long-term disability. It is estimated that 5.8–6.5 million people currently live with stroke related disability in the US and that this number will increase by 20.5% by 2030. The current US health infrastructure is not prepared for these increasing numbers. Limitations in health insurance coverage and the shortage of rehabilitation practitioners decrease access to rehabilitation. Community-based settings are becoming viable venues for delivering long-term post-stroke care, however, they are plagued by staff with limited expertise, low number of therapists and lack of financial resources for rehabilitation. Because of this, the quality of care is compromised, and functional outcomes of patients are not equal to hospital-based rehabilitation settings. We seek to develop a novel solution to this problem. Implementing affordable design is a fundamental strategy for increasing access to rehabilitation technology for patients regardless of socio-economic status. Doing so, decreases healthcare disparities and reduces long-term healthcare costs. We propose to use affordable robots to improve access to quality rehabilitation care in low-resource, community-based settings. In Phase 1, we leverage a 1 degree of freedom haptic robot with control algorithms to develop a beta version of the robot hardware and software. The new robot have a novel end-effector to allow more diverse arm and hand exercises, be connected to cloud-based gaming, and provide patient-specific therapy that adjusts for motor impairment and cognitive impairment. 15 stroke patients with a wide range of motor impairment levels will complete clinical assessments of motor and cognitive impairment followed by robot-based assessment and therapy games. Subjects will be instrumented with sensors monitoring key upper extremity muscle activity, trunk activity and heart rate during robot tasks. A key milestone will be to identify kinematic metrics from the robot tasks that strongly correlate and predict clinical scores of motor and cognitive impairment. Another milestone will to drive patient-specific strategies by adjusting the robot’s control parameters and the game parameters. In Phase 2, we will develop the hardware to allow three haptic robots to dock (a gym) and be configured to allow patients to play therapy games alone or collaboratively. We will test the safety and feasibility of the gym in a community-based rehabilitation setting where stroke patients typically receive 1 hour each of physical therapy (PT), occupational therapy (OT) and speech therapy (SLP). 36 patients will be randomized to either a robot (RT) or a control group (CT). Both groups will receive PT and SLP, but the RT will receive the robot gym therapy targeting the upper limb and the CT will receive a dose-matched hour of OT. Therapy will occur over 4 weeks with two follow-up assessments. Key milestones will be to show that the RT has the same or better functional outcom...