PROJECT SUMMARY Aging is a terminal process that affects all biological systems. Biological aging—in contrast to chronological aging—occurs at different rates for different individuals. In humans, growing old comes with increased health issues and mortality rates, yet some individuals live long and healthy lives, and others succumb earlier to diseases and disorders. The concept of frailty is used to quantify this heterogeneity and is defined as the state of increased vulnerability to adverse health outcomes. The frailty index (FI) is an invaluable and widely used tool which outperforms other methods to quantify frailty. FIs have been adapted for use in mice using a variety of both behavioral and physiological measures as index items. However, because conducting mouse FI requires trained individuals for manual scoring, it often limits the scalability of the tool. Thus, although the FI is an extremely useful tool for aging research, an increase in its scalability, reliability, and reproducibility through automation would enhance its utility. We used machine learning applied to video data to create an automated visual FI (vFI). The is easy to implement, unbiased, and scalable. Here we propose to improve our tool and carry out an interventional study. We will adopt the vFI to function with genetically diverse mice (R61: Aim 1). We will also create features from long-term monitoring to increase accuracy and breadth of systems measured in the vFI (R61: Aim 2). Finally, we will apply the vFI to a diet intervention study to show its utility for large scale studies (R33: Aim 3). We will test a high fat high sugar diet (increased frailty) and caloric restriction group (decreased frailty) with normal chow (control) in a Diversity Outbred population of mice. The result of this project will be a fully validated and automated vFI that can be used for high-throughput interventional studies, enabling therapeutics for healthy aging.