Falls are the leading cause of fatal and non-fatal injuries in persons older than 65. Injuries due to falling result in 2.8 million emergency visits annually, and 25% of falls result in severe injuries (such as fractures or traumatic brain injury). Research has shown that vibrotactile sensor augmentation can alert a wearer to the possibility of a fall and reduce the probability of a fall. We propose a wearable system consisting of a camera, inertial measurement unit, and on-board computing that can observe the environment in front of the human, along with their current gait, and utilize machine learning methods to predict with high certainty their expected future path and gait over the next few steps and leverage this information to alert the wearer via vibrotactile response if the risk of falling based on the path is probable.