Among adults aged 65 years and older, falls are a serious, growing, and costly public health problem. Falling can lead to reduced mobility, functional decline, and loss of independence. Detecting and mitigating fall risks are effective for avoiding the serious consequences of falls. Smart insoles have the potential to improve upon current clinical fall risk assessments by taking less time and by enabling automatic assessments during a daily routine. However, existing smart insole systems face two major hurdles. First, the flexible pressure sensor arrays commonly used in them are prone to errors that slowly change over time, called “sensor drift.” Sensor drift makes it hard to accurately measure critical fall risk assessment values, such as gait, balance, and leg strength. Second, the artificial intelligence models that are widely used to assess fall risk lack clinically meaningful explanations of the results, which are needed to identify fall risk factors, inform effective interventions, and build trust. This project will develop and evaluate a new auto-calibrated insole system along with explainable artificial intelligence (XAI) models. The project will enable accurate, tailored, and reliable fall risk assessment during the normal daily routines of older adults. This research will result in fall risk assessment technologies that promote health, independence, and overall quality of life for older adults. This project also provides education and training for K-12, undergradu