SPECIFIC AIMS__________________________________________________________________________ Today, nearly 37 million people are living with HIV (PLWH) worldwide, and 30 to 40% of them have motor and cognitive impairment due to HIV-associated neurocognitive disorders (HAND) and HIV-dementia, even with antiretroviral therapy (ART). Further, PLWH face an increased (1.5-fold) risk of stroke due to HIV vasculopathy and other comorbidities, making it an HIV-associated non-communicable disease. These cognitive and motor disabilities impede activities of daily living (ADLs) and reduce quality of life. Unfortunately, the burden of HIV and the consequence of cognitive and motor impairments disproportionately affect PLWH in low-and middle- income countries (LMICs), particularly in Sub-Saharan Africa (~25 million). Botswana, as both a country with the fourth largest HIV prevalence in the world (20.3% in 2017 about 370,000 of 2 million persons) and an extensive HIV treatment program, is an ideal setting for our proposed study. It was the first Sub-Saharan African country to rollout free universal ART programs. Regional studies in Africa have consistently reported high prevalence of neurocognitive impairment in PLWH despite ART. A systematic review and meta-analysis of HAND in subSaharan Africa, cognitive impairment estimates in Uganda were 46.49% (pre-ART) and 28.50% (post-ART). Since motor and cognitive impairments in general and HAND in particular go undiagnosed in PLWH due to extremely scare infrastructure and healthcare resources, novel strategies are needed to offer both evaluation and treatment. Our long-term goal is to develop and use cost-effective robot and computerized technologies to bridge rehabilitation care gaps in LMICs and leverage these innovative tools to help early and objectively assess, diagnose and ultimately treat PLWH presenting with motor and cognitive impairment due to NCDs. The rehabilitation robotics field, which has developed successful low-cost neurorehabilitation strategies for stroke and other neurological injuries, provides a usable template to address this challenge. We have created a series of low-cost robotic systems with 1 degree of freedom haptic robot prototypes for the upper arm that uses motor and cognitive exergames for assessment and therapy. Requiring little human supervision, preliminarily results suggest the potential to evaluate cognitive and motor function with robust validity compared to intensive testing by trained assessors in PLWH and Stroke Survivors in the USA. We have recently deployed two of these systems (Rehab CARES) each with two haptic Theradrive robots in Botswana. We successfully trained local engineers to assemble, test, and repair the robots and manufacture the robot mounting bases. We have also trained local clinicians and engineer researchers to administer clinical motor and cognitive metrics and use the robots with patients with HIV, stroke, and HIV-stroke. Results of this preliminary work in Botswana support our ability 1) to recruit the target population for this grant; 2) implement and maintain usable and affordable robot devices in two hospital environments; 3) administer the robot assessment and derive the objective robot-based metrics; and 4) correlate the robot-based metrics to clinical metrics for motor impairment and cognitive impairment. Lessons learned lead us to this R21 proposal. Here, we would like to establish concurrent validity of robot-based metrics developed, implement an affordable tech-based gait assessment and metrics, perform test-retest reliability of all technology systems and finally, determine how feasible it would be to use them to classify HAND. To do so, we leverage the new Botswana Rehabilitation and Robotics Lab that contains three labs: the robot lab, the motor lab and the cognitive lab. We also leverage the help and guidance of BotswanaUPENN Partnership (BUP), a 21-year-old successful collaboration between the University of Pennsylvania, the University of Botswana, and the Ministry of Health and Wellness in Botswana. Our R21 specific aims are:                                     Aim 1: To determine concurrent validity and test-retest reliability of an affordable robot to objectively ASSESS cognitive impairment and motor impairments in PLWH. 40 PLWH--with and without stroke--and 40 HIV-negative controls will be evaluated using both clinical metrics and robot/tech-based metrics at two time points, at least 1- week apart. We hypothesize that robot/tech-based metrics will have test-retest reliability as well as concurrent validity with individual clinical measures of motor and cognitive function that usually require rehabilitation expertise and manpower to complete. Since some of the metrics have not been validated on HIV patients for Botswana’s context, the HIV controls are being used to verify established norms. Aim 2: To determine the feasibility of robot/tech-based classification of HIV-associated neurocognitive disorders (HAND). 80 PLWH without stroke will be evaluated using standardized neurological and neuropsychological (NP) examination required to meet minimal HAND Frascati criteria along with the robot/tech measures. We will use the standardized criteria and Global Deficit Scoring to classify the participants into asymptomatic neurocognitive impairment (ANI), mild neurocognitive disorder (MND) and HIV-associated dementia (HAD). We will use the metrics derived from our assessment games to classify PLWH similarly to the Frascati Criteria. We hypothesize that the technology-based metrics will classify HAND with greater sensitivity and specificity when compared with HAND screening tools. The Receiver Operator Curve (ROC) will be used as empirical validation. We will track known predictors of HAND such as age, sex, education, viral load, gait, substance abuse, and tobacco use to assess their impact on models. These aims support the use of low-cost robots for objective assessment and potential treatment of neurocognitive-motor impairment in PLWH in LMICs.