Alzheimer’s disease (AD) affects over 5 million older adults in the US. AD is associated with accelerated deterioration in physical function, resulting in a higher risks of falls, decline in mobility, reduced quality of life, and loss of independence. Exercise interventions for individuals with AD or mild cognitive impairment (MCI) have shown promising improvements to overall health, physical function, and independence. Unfortunately, routine exercise is difficult to maintain for these individuals due to safety concerns and the need for supervision and guidance. To address this need, we developed an exercise guidance app for our touchscreen Flint Tablet that connects to a sensorized, low-profile stationary cycle (Cycli). We evaluated the feasibility of this system in promoting home exercise in individuals with MCI or mild dementia in a Phase I SBIR. The exercise guidance app and overall approach received positive feedback. However, the Cycli hardware had a number of critical issues that prevent it from being a viable home exercise option for this population (e.g. it created a potential tripping hazard and was not engaging for some users). To address these limitations, we propose in Phase II to integrate the guided exercise app developed in Phase I with our existing FitMi system, which is a wireless, sensorized at-home exercise device originally developed for individuals with stroke. We refer to this new system here as FitMi AD. The existing FitMi hardware includes a patented, multipurpose wireless input device (a “Puck”) and a wearable wireless input device (a “Clip”) that measure both movement and compression force. Thus, we believe that with appropriate, clinically informed design improvements, FitMi AD will facilitate home exercise in individuals with MCI or mild dementia due to AD. Specifically, in addition to incorporating the scheduling, tracking, and goal-setting features developed and validated in Phase I, FitMi AD will include a new library of exercises from Exercise: A Guide from the National Institute on Aging (NIA). FitMi AD will also use real- time data from a Bluetooth heart rate monitor to automatically adjust the exercise intensity to achieve a beneficial level of exertion. Our specific aims for this Phase II project are to: Aim 1) Perform structured interviews with clinical experts and individuals with MCI or mild dementia due to AD and their caregivers to optimize the design of FitMi AD (n=25); Aim 2) Develop the FitMi AD exercise software with iterative expert and user feedback; Aim 3) Determine the safety and effectiveness of home-based exercise training with FitMi AD in a randomized controlled trial with individuals with MCI or mild dementia due to AD (n=30). Our hypothesis is that participants who exercise at home with FitMi AD will show significantly greater improvements in physical function than the control group. If successful, FitMi AD will be a safe, effective, and commercially viable home exercise tool for individuals ...